Marler Clark has been retained by one of the Washington residents sickened in the Salmonella Bovismorbificans outbreak linked to broccoli sprouts grown by Evergreen Fresh Sprouts, LLC of Moyie Springs, Idaho. A disclosure, as always: the firm sued Evergreen on behalf of four people in its 2014 E. coli outbreak, and I have represented people sickened by raw sprouts since the early 2000s. Read accordingly.

Thirty-two sick in six states, and the federal pages have not moved since September 24

CDC’s outbreak page and FDA’s advisory were last updated September 24. They count 32 people infected with the outbreak strain in six states: 24 in Washington, four in Montana, and one each in Idaho, Minnesota, Oregon and Utah. Three have been hospitalized. No one has died. Illnesses began between July 7 and September 3, 2026. Patients range in age from 2 to 87, with a median of 52. Of the 27 people interviewed, 26 reported eating sprouts or microgreens; 22 named broccoli sprouts, and one a microgreen mix containing broccoli microgreens. CDC measures that against the 8.7 percent of people in its FoodNet Population Survey who report eating any sprouts, and concludes that broccoli sprouts are what made these people sick.

CDC also says the true number is likely much higher, because most people with Salmonella are never tested and because it takes three to four weeks to tie a new illness to an outbreak. Eleven days have passed without a federal update. CDC’s page still reads “Investigation status: Open.”

The Washington State Department of Health updated the same day: 24 cases in nine counties and two hospitalizations.

CountyCases
King7
Whatcom5
Spokane4
Clark2
Skagit2
Island1
Snohomish1
Walla Walla1
Yakima1
Total24

Washington cases by county of residence as of September 24, 2026 (7 + 5 + 4 + 2 + 2 + 1 + 1 + 1 + 1 = 24). Source: Washington State Department of Health.

This outbreak is separate from the alfalfa sprout outbreak traced to Everything Sprouts of Minnesota, which CDC put at 76 sick in 16 states on October 1. Between the two, 108 people have been sickened by raw sprouts this summer.

How the sprouts were tied to the illnesses

Whole genome sequencing showed the bacteria from the 32 patients are closely related, which is how PulseNet found the cluster. Then the product matched, three times. Montana investigators collected an unopened package of Evergreen broccoli sprouts from a sick person’s home, and sequencing showed it carried the outbreak strain. Washington investigators collected two more samples, one leftover from a patient and one from a retail location, and both carried it. FDA’s and the states’ traceback identified Evergreen Fresh Sprouts as the grower. In the company’s three outbreaks since 2011, this is the first time a product sample has matched the people.

Thirty of the 32 patient isolates and two food samples showed no predicted antimicrobial resistance. Two patient isolates were predicted resistant to streptomycin, sulfisoxazole and tetracycline, which CDC says is unlikely to affect the antibiotics used to treat most patients.

The recall came after the illnesses

FDA says Evergreen initiated a recall on September 4, the day it stopped producing broccoli sprouts, after the Montana sample tested positive. The company’s notice reached FDA’s recall page on September 10. It covers 215 cases, six four-ounce bags to a case, delivered to three Washington distributors between August 24 and September 2, with use-by dates of September 7, 9, 11, 14 and 16 and UPC 8 38796 00105 1. FDA says the sprouts were sold to grocery stores, restaurants and food service customers in Idaho, Montana and Washington and may have been distributed further. Minnesota, Oregon and Utah each have a patient, and none of the three is on the distribution map.

Set the recall against the illnesses. The first people got sick on July 7. As I wrote on September 24, 25 of the 32 onsets fall on or before August 23, the day before the first recalled bag was delivered. The recall could have reached, at most, the last seven illnesses. For the other 25, the sprouts were eaten and gone weeks before anyone said the product had a name. That is the ordinary shape of a sprout outbreak, and it is why a recall is a public health tool and not a measure of who was hurt.

Salmonella has been found in a broccoli seed lot, and FDA’s outbreak pages do not mention it

On October 2, FDA posted a recall by EAK Distribution LLC of Casper, Wyoming, of HOME GROWN Green Sprouting Calabrese Broccoli Seeds sold on Amazon to customers nationwide between August 20 and September 18, lot code AUG172026, UPC 850056627443. EAK says its seed supplier told it on September 16 that bulk broccoli seed lot W4105, the lot EAK had packaged, had been “identified as being used by a sprout grower whose finished product were part of an FDA investigation” into a multistate outbreak; that the supplier recalled the bulk lot; and that on September 28 the supplier reported FDA had collected and analyzed a sample of the seed and detected Salmonella.

The notice names neither the supplier, the grower nor the outbreak, and it does not have to. FDA has two sprout outbreaks open. One is alfalfa, traced to a Tennessee seed lot. The other is broccoli sprouts grown by Evergreen. A broccoli seed lot used by a sprout grower whose product is in a multistate outbreak investigation, with Salmonella in the seed, describes this outbreak. FDA has not said so. Its Evergreen advisory still says the agency is working “to determine the source of contamination,” and its outbreak table records that sampling has begun and reports no result. The supplier’s recall of the bulk lot, which EAK says happened by September 16, is not on FDA’s recall page. I wrote about the notice in detail on Friday.

Seed does not expire the way sprouts do; a pouch bought in August will be in a pantry next spring. It is the second time in six weeks that the seed behind a sprout outbreak has reached home sprouters. In the alfalfa outbreak the implicated lot was sold by Todd’s Seeds and by Berlin Seeds, into 44 states.

The third Evergreen outbreak since 2011

I set out the company’s record on September 3. In 2011, CDC linked 25 Salmonella Enteritidis illnesses in five states to alfalfa and spicy sprouts sold under the Evergreen Produce label; the company declined FDA’s request for a recall until the agency warned the public directly, then recalled on July 1. In 2014, 19 people in six states were infected with E. coli O121 from Evergreen’s raw clover sprouts and seven were hospitalized; there was no recall, and FDA’s inspection report described condensate and irrigation water dripping from rusty valves and a corroded pitchfork used to move sprouts. Marler Clark filed four lawsuits on behalf of people sickened in that outbreak. With this year’s 32, the company’s three outbreaks account for 76 illnesses.

What the law says for the people who got sick

Washington’s Product Liability Act makes a manufacturer strictly liable for harm caused by a product that was not reasonably safe in construction. The claimant does not have to prove negligence. Fresh sprouts carrying Salmonella are a product that was not reasonably safe when it left the grower, and, depending on the facts, the stores and restaurants that sold or served them may be liable as well. The proof is the proof the health departments already assembled: a laboratory-confirmed infection, a whole genome sequencing match to the outbreak strain, and evidence of what was eaten: a receipt, a loyalty card record, a restaurant order or the health department’s own interview. Damages include medical bills, lost wages and the cost of recovery, and Salmonella leaves some people with reactive arthritis or irritable bowel syndrome that does not resolve.

If you were sickened in this outbreak, keep the package or a photograph of it and any receipt, report your illness to your health department (Washington’s online form is here), and ask that your stool isolate be preserved. Isolates are routinely discarded, and sequencing is what ties an illness to the outbreak.

A label before the next one

FDA has told children, older adults, pregnant women and people with weakened immune systems not to eat raw sprouts since 1998, and that advice has never appeared on a package of sprouts or a pouch of sprouting seed. The draft citizen petition asking FDA to require a warning statement on both is open for comment through October 15; the revised draft is here. Read it and tell me what I got wrong before I file it.

Marler Clark represents a Washington resident sickened in this outbreak and continues to investigate it. The firm’s outbreak page is here. Call 1-800-884-9840 or email bmarler@marlerclark.com.

On October 2, 2026, FDA closed its investigation of the largest multistate Cyclospora outbreak on record. FDA said the two positive environmental samples – one from a tank of outgoing wastewater at the Taylor Farms de Mexico processing plant, one from a drainage ditch at a grower in the traceback – add to the strong epidemiological data and traceback convergence supporting the link to the recalled iceberg lettuce.

No lettuce tested positive. Taylor Farms has made that its point. Its Cyclospora information hub stresses the absence of a positive product sample, even as it says it stands behind the decision to recall.

That argument has surface appeal. It borrows the vocabulary of laboratory proof. And it asks for a standard that no public health agency in this country uses.

What an outbreak link is built on

Investigators use three kinds of evidence: epidemiology, traceback, and food or environmental testing. CDC’s published investigation steps say that when food eaten by several sick people shares a common point in the supply chain, that convergence can confirm the food is the source. Finding the strain in a food can “also” confirm it. Also – not only.

CDC goes further. Officials can act on strong epidemiological evidence alone and do not have to wait on a laboratory to find the outbreak strain in the food. And CDC says the reverse is true too: finding the strain in a food cannot by itself confirm the source unless interviews show sick people ate it.

CDC’s Field Epidemiology Manual puts it in training terms. Investigators generally identify the likely source when two of the three lines of evidence are clear and convincing. For short shelf-life foods – the manual uses leafy greens as its example – testing may be impossible, so investigators rely on the other two. And a negative result does not rule a food out.

USDA’s Food Safety and Inspection Service defines an implicated product the same way: strong evidence from two of the three sources. The CIFOR guidelines, written by the state epidemiologists, local health departments and public health labs with CDC, FDA and USDA, tell investigators to pull a food when exposure data link it to illness even if the pathogen has not been isolated from the food.

Why the lettuce is never there to test

The lettuce that made people sick was eaten three or four weeks before anyone knew there was an outbreak. That is CDC’s own reporting lag. What gets tested is what is left – which is, by definition, other lettuce.

And contamination is sparse and uneven. Do the arithmetic. If one head in a hundred is contaminated and you test sixty with a perfect test, every result comes back negative 55 percent of the time. At one in a thousand, 94 percent. For Cyclospora, which cannot be cultured and is hard to detect on produce at all, the test is far from perfect.

In FDA’s fall 2019 romaine investigation, teams took 406 sub-samples across Salinas Valley farms. One matched. The only product positives came from two intact packages sitting in sick people’s refrigerators. That is how product positives happen – luck, not systematic testing.

The industry has said this itself

Here is where the defense runs into its own allies. Western Growers told growers this July that testing for Cyclospora is not recommended because of sampling and technological limits that may give a false sense of security. The California and Arizona Leafy Greens Marketing Agreements opened their 2026 webinar with “Testing is not the answer.”

The International Fresh Produce Association’s own 2026 outbreak FAQ says the recall and the outbreak findings rested on epidemiology and traceback, not on a positive lab test, and that the absence of a positive sample did not mean those findings were wrong.

In November 2018, with no grower, processor or brand identified and no romaine positive, the produce industry urged a halt to all romaine shipments. That was the industry acting on epidemiology alone. It was right to.

The record

The 2018 Yuma romaine outbreak – 210 sick, five dead – was solved without a single positive head of romaine; the strain turned up in canal water. In 2013, Nebraska and Iowa traced a Cyclospora outbreak to salad mix from Taylor Farms de Mexico on interviews and records, with no product positive. Taylor Farms suspended production on that evidence. Same company, same kind of proof, thirteen years ago.

In 2011, Germany stopped an E. coli O104:H4 outbreak that sickened more than 3,800 people by implicating sprouts on a case-control study, a restaurant cohort and product tracing. As a 2025 Eurosurveillance perspective put it, laboratory evidence is neither sufficient on its own nor a prerequisite for action.

Closer to home, after a negative lettuce test in the November 2024 St. Louis outbreak, the county health department said plainly that a negative result did not rule out the salad. The epidemiology was the most compelling evidence.

What the argument really asks for

Take the “nothing tested positive” defense seriously and follow it to the end. The implicated product is almost always gone. Produce testing is insensitive. Some pathogens cannot be cultured at all. A rule that no company is the source until its product tests positive is a rule under which no produce outbreak could ever be attributed to anyone.

That is not a standard of proof. It is a standard of immunity. CDC, FDA, USDA, CIFOR, Western Growers, the LGMAs and IFPA have each rejected it in writing.

I spent four hours Sunday morning taking the deposition of an expert retained by the defense in the E. coli O157:H7 outbreak of November 2024 – the one traced to romaine lettuce that sickened 89 people in 15 states, hospitalized 36, put seven into kidney failure and killed one, and that FDA closed in January 2025 without a word to the public and without naming the grower or the processor. Four of the seven with HUS are my clients; two are children who each spent about two weeks on dialysis. The expert’s opinion was that children dialyzed that long have a less than fifty percent chance of needing a kidney transplant. That is the defense: the odds are better than a coin flip. It is little solace to a child who has already had a catheter in a vein, or to the parents who sat beside the machine. It may explain why I am less patient than usual today with the leafy greens industry, and with one grower and processor in particular. The company my clients are suing over that romaine, which denies its product was the source, is the company whose outgoing wastewater FDA reported positive for Cyclospora on Friday.

A disclosure, as always: Marler Clark represents people sickened in this summer’s outbreak and in the 2024 romaine outbreak and has filed suits against the Taylor Farms companies in both. Read accordingly.

On October 2, FDA closed its investigation of the largest Cyclospora outbreak in American history – 12,883 sick, 570 hospitalized and two dead in 21 states – and reported the parasite in outgoing wastewater at the Taylor Farms de Mexico plant and in a drainage ditch at a traceback grower. Taylor Farms rewrote its outbreak page the same day. Mexico’s Secretaría de Salud and its regulator, Cofepris, answered on October 3. Guanajuato’s Secretaría del Campo issued a bulletin that afternoon. Side by side, the first two could have come from one outline.

The argument, in two voices

Taylor Farms says no product sample tested positive, that its own roughly 2,000 samples in central Mexico since May were all negative, and that epidemiology cannot confirm a product caused illness without a positive lab test of the product. Cofepris says its National Reference Laboratory tested raw material, finished product and water from the plant and found nothing, and that FDA’s environmental findings do not, by themselves, constitute evidence that agricultural products were the source.

Half of FDA’s sentence. Both quote FDA’s line that there is not enough evidence to conclusively determine how contamination occurred; Taylor Farms repeats it twice, Cofepris builds on it, and neither finishes the sentence.

The samples prove nothing. Taylor Farms says the wastewater sample was discharged sewer water outside the facility, taken more than three weeks after production stopped, and that the grower with the positive ditch is no longer a supplier. Cofepris says finding organisms in waste drains does not by itself show that vegetables bound for export were contaminated.

Each draws the same conclusion. Taylor Farms: FDA has no confirmed positive for any Taylor Farms product. Cofepris: the available information does not establish, through conclusive analytical evidence, that Mexican products were the origin of the outbreak – and said the same of the jalapeño Salmonella Javiana outbreak closed the same day.

What the argument leaves out

The other half of FDA’s sentence. FDA’s October 2 update reads in full: 

“While there is currently not enough evidence to conclusively determine how contamination occurred in this outbreak, the positive samples show that Cyclospora was present in the environment where lettuce was grown and processed. This adds to the strong epidemiological data and traceback convergence supporting the link between this outbreak and recalled shredded iceberg lettuce from Taylor Farms de Mexico.” 

The same update says the traceback identified several farms and a processing facility and that FDA remains confident the recalled lettuce is off the market. That is an agency standing behind its conclusion. The half-sentence traveled to Salinas and Mexico City; the whole one did not.

Mexico’s negatives have a date. Cofepris collected ten iceberg samples and four water samples on July 18 through 20. The recalled lettuce shipped from June 29 to July 16, and the plant suspended production on July 18. IFPA’s own page explains why that matters: implicated product is often gone by the time testing becomes possible. A negative on July 20 says nothing about June 29.

Taylor Farms’ 2,000 samples have a question attached. The company told the New York Times in August that its fecal indicator testing covered all its water sources and every test was negative. Indicator tests look for bacteria; FDA’s validated method looks for the parasite and found it. The company has not said what its 2,000 samples tested for.

The genotype answer has two halves. Taylor Farms’ FAQ asks whether the environmental positives match the outbreak genotype and answers No, then concedes that FDA is still analyzing the results and had not confirmed a match as of October 2. Not yet confirmed and No are different answers.

The industry already answered this argument, in writing

IFPA’s consumer page, frozen on September 14, answers the question Was a specific food identified? with Yes. It says the recall and the outbreak findings rested on epidemiology and traceback, not a positive laboratory test, and that the absence of a positive sample did not mean the epidemiological and traceback findings were wrong. Western Growers told its members in July that testing for Cyclospora is not recommended, because sampling and technological limits make it of little value and may provide a false sense of security. The California LGMA’s statement accepted Mexican-grown lettuce as the likely source.

Those are the industry’s words, and they describe what two thousand-odd negatives are being used for now. On Friday I asked the industry to retire the water story; the replacement is the no-product-positive story, now with a government behind it. Since October 2, I can find nothing from IFPA, Western Growers or the leafy greens marketing agreements on FDA’s findings. Taylor Farms co-founded the Center for Produce Safety, helped create the LGMAs and hosts the Western Growers Center for Innovation and Technology. No association wants to referee between a founding member and the FDA.

But the silence has a cost. In Mexico, the only argument readers have seen is Cofepris’s. In Washington, a standard is hardening – no product positive, no case – that will be used against the next investigation, including the next one that clears California growers.

A request to eight people

Yesterday, I published a list of what would fix this and offered the industry three draft citizen petitions. Today I am asking eight people, by name, to answer. Max Teplitski and Eric Stevens at the International Fresh Produce Association. De Ann Davis and Joelle Mosso at Western Growers. Tim York at the California Leafy Greens Marketing Agreement. Hilary Thesmar at FMI. Greg Ferrara at the National Grocers Association. Lance Jungmeyer at the Fresh Produce Association of the Americas, which represents Mexican produce importers.

Two things.

First, go down the list and say, item by item, what you support, what you oppose, and why. The items are short: repeal the traceability delay and comply with FSMA 204 now; numeric standards and routine testing for pre-harvest water on high-risk produce; parasite testing of growing and processing water where outbreaks have happened; processor wastewater kept away from fields, canals and product; a kill step that works on parasites; feedlot setbacks and watershed monitoring; the same standards for imports under FSVP; Cyclospora back in FoodNet; a produce testing trip wire like the Microbiological Data Program; one genotyping standard; data shared with CDC and FDA in hours; verified worker sanitation at supplier farms; published root cause analyses, Taylor Farms’ water results included; and no more support for FSMA rollbacks. If an item is wrong, say why; if it works only in another form, say which.

Second, ask me for the three petition drafts: FDA sampling at feedlots implicated in outbreaks, naming the firms implicated in outbreaks, and consignee lists with ten-day classification for every Class I recall. Tell me what is wrong with them; I would rather hear it from you than from FDA.

I will publish every response in full, unedited and under the author’s name, here, and I will say where it changed my mind.

The Global Produce & Floral Show opens in Orlando on October 13. An answer before then would let the industry walk in with a position on the record. My email is bmarler@marlerclark.com.

One plant, one grower

FDA’s finding is narrow: one plant and one grower in Guanajuato. An argument that says that finding proves nothing is an argument that no finding will ever prove anything about this parasite, because no product will ever test positive in time. The industry can say that, or it can say what its own page said on September 14. It cannot say both. I would like to know which.

In September 2006, bagged baby spinach grown in California’s Salinas Valley sickened over 200 people with E. coli O157:H7 and killed five of them. For a few weeks you could not give spinach away. Twenty years later, almost to the month, FDA closed its investigation into the largest Cyclospora outbreak in American history: 12,883 confirmed illnesses, 570 hospitalizations and two deaths in 21 states, linked to shredded iceberg lettuce from Taylor Farms de Mexico. Inspectors found the parasite in a wastewater tank at the processing plant and in a drainage ditch at a grower, and FDA still cannot say how it got onto the lettuce.

Between those two outbreaks sit two decades of the produce industry deciding, repeatedly, how much food safety it was willing to pay for. Some of those decisions were genuinely good. Some were not. In July I wrote Reap What You Sow, about how the produce lobby helped kill the only federal program testing produce for pathogens, and I am not going to repeat all of that here. What I want to do is be fair about the whole record and then offer the industry a list of things we should be able to agree on. The International Fresh Produce Association meets in Orlando October 13–16 – I must have misplaced my invitation?

 Consider this an agenda without me being there.

A disclosure, as always: Marler Clark represents people sickened in this summer’s Cyclospora outbreak and has filed lawsuits against the Taylor Farms companies. I have been suing produce companies for more than thirty years. Read accordingly.

Who we are talking about

The industry consolidated a few years ago. The Produce Marketing Association and United Fresh merged in 2022 to form the International Fresh Produce Association (IFPA), which is now the dominant voice in Washington. Western Growers represents the farmers in California, Arizona, Colorado and New Mexico who grow most of the country’s lettuce and leafy greens, and it writes the food safety metrics for the California and Arizona Leafy Greens Marketing Agreements (LGMA). On the buy side, FMI – The Food Industry Association and the National Grocers Association speak for the retailers. Behind them are dozens of commodities and regional groups, from the National Onion Association to the Fresh Produce Association of the Americas in Nogales, which represents the importers of Mexican produce.

What they got right

After spinach, the leafy greens industry did something it had never done before. Working through Western Growers, it wrote commodity-specific food safety metrics for growing and harvesting leafy greens, and in 2007 California created the LGMA to put signatories under government audit. It is not perfect. I have sued LGMA members. But it was a real system where there had been none, and after the Yuma and Salinas romaine outbreaks of 2018 through 2020, the LGMA tightened its water and harvest requirements.

The industry also paid for science. PMA put up the seed money for the Center for Produce Safety at UC Davis, which has funded applied produce safety research for nearly two decades. After the 2008 Salmonella Saintpaul fiasco, when tomatoes were blamed for weeks before jalapeños turned out to be the problem, the industry launched the Produce Traceability Initiative. It understood then that the inability to trace is what turns one bad farm into a whole category’s catastrophe. United Fresh supported passage of the Food Safety Modernization Act in 2010.

And this summer, some of the industry said the right things. IFPA called on Congress to strengthen FDA’s Human Foods Program, implement end-to-end traceability and fund state produce safety programs, and it said its members are on the record against continued delays of the Traceability Rule. Taylor Farms posted that it supports traceability legislation. Western Growers CEO Dave Puglia said the industry needs to open its doors wider and show consumers its food safety practices. All of that is good. Now the industry must act on it.

What they got wrong

They killed the trip wire. In 2012 the USDA’s Microbiological Data Program, about $5 million a year to sample produce in commerce for pathogens, was zeroed out. It had triggered 23 produce recalls in 2010 and 2011 alone. Dan Flynn laid the kill at the feet of United Fresh’s lobbyists. Thirteen years later, on July 1, 2025, CDC dropped Cyclospora from FoodNet surveillance. When this summer’s outbreak arrived, nobody was counting.

They fought water standards. The 2015 Produce Safety Rule set numeric limits for generic E. coli in irrigation water. The industry called them unworkable, FDA delayed them, and in 2024 FDA replaced numeric testing with “systems-based” assessments in which testing is required only in certain circumstances. Today no law requires a farm to test its irrigation water or its produce for Cyclospora, and the antimicrobial washes the LGMA requires are validated against generic E. coli, not a parasite that chlorine does not kill.

They cheered the traceability delay. FDA’s Food Traceability Rule was supposed to take effect January 20, 2026. FMI called it overly complex, and when FDA pushed compliance to July 20, 2028, FMI and NGAapplauded. Congress then wrote the delay into the November 2025 appropriations act. IFPA says it opposed the delays, and I take that at face value. But its own 2025 Impact Report touts supporting FDA’s “10 to 1” deregulation effort to reduce FSMA-related burdens. You cannot be for end-to-end traceability and for trimming FSMA at the same time.

They argued with the epidemiologists. On July 15, with cases climbing, IFPA’s chief science officer publicly criticized Michigan health officials for discounting recreational water as a risk factor. The next day CDC and FDA named the lettuce. In August, IFPA filed comments asking CDC to look harder at recreational, occupational and water exposures. Wanting good epidemiology is fine. Floating alternative theories in the middle of the largest outbreak of its kind is not, and it spent credibility the industry will need the next time.

They studied the neighbors instead of regulating them. The 2018 Yuma romaine outbreak was tied to irrigation canal water near a large cattle feedlot. Eight years of studies later, there are still no enforceable setbacks between feedlots and the fields and canals that serve leafy greens.

They took a victory lap on testing. IFPA claims credit for prompting FDA to change its Cyclosporadetection method to cut false positives. Maybe the change was warranted. But when the only laboratory tool you have for a parasite is already weak, any change that produces fewer positives deserves independent scrutiny, not a press release.

Here is the cost of all of that. Puglia told the New York Times that lettuce sales fell near 30 percent this summer, and millions of heads of healthy California lettuce were plowed under. California grew none of the outbreak lettuce. That is what happens when you cannot trace a product in a day: the innocent pay for the guilty. The industry has more reason than I do to fix this.

Where we can find common ground

I said in July that I would rather work with these organizations than shout at them. I meant it. So here are two big asks, a list, and three petitions.

First, ask Congress to unwind the traceability delay. The Safe Food Coalition asked Congress in August to rescind it. IFPA says it supports end-to-end traceability. Taylor Farms says it supports traceability legislation. What stands in the way is a congressional directive and the retailers who lobbied for it. IFPA, Western Growers and the LGMAs should put their names on a letter asking Congress to repeal the enforcement bar and let FDA enforce FSMA 204 on a near-term date, ideally before the 2027 growing season for leafy greens and the other foods on the Food Traceability List. And whatever Congress does, members who grow, pack and buy those foods should comply now. Western Growers has already written its members a guidance document on how. The work is done; the only thing missing is the will.

Second, push forward on water. Cyclospora is a water problem. Both of FDA’s positive samples from Mexico were water. The industry should ask Congress and FDA for the following:

  1. Enforceable, numeric microbial standards for pre-harvest water used on high-risk produce eaten raw, with routine testing. The LGMA already tests; make that the national floor, not the ceiling.
  2. Cyclospora and other parasites added to the testing of growing and processing water at operations producing Food Traceability List commodities, starting in regions with an outbreak history.
  3. Processor wastewater management that keeps outgoing water away from fields, canals and product. FDA found the parasite in a Taylor Farms de Mexico wastewater tank. In 2013, after an earlier Cyclospora outbreak, Taylor Farms said it was testing water at that operation. I have never found a public result.
  4. Research money for a validated kill step or water treatment that works against parasites, including restoring the USDA Cyclospora research projects that were shelved this year.
  5. Enforceable setbacks and watershed monitoring for concentrated animal feeding operations near produce fields and irrigation sources.
  6. The same standards for imports, enforced through the Foreign Supplier Verification Program, which FDA says it will expand before the 2027 season.

And then the rest of the list. None of these should be controversial:

  • Restore Cyclospora and the other dropped pathogens to FoodNet and fund the state and local labs that find outbreaks.
  • Rebuild a produce testing trip wire like the Microbiological Data Program. At roughly $5 million a year, it is the cheapest insurance policy this industry will ever be offered.
  • Fund and support the single Cyclospora genotyping standard FDA and CDC say they are building. IFPA itself has said the genomic tools for this parasite lag far behind those for bacteria. Here is a chance to fix it together.
  • Agree now, before the next outbreak, on a protocol for sharing shipping, sourcing and test data with CDC and FDA within hours. Western Growers’ GreenLink platform already holds years of industry test data; share it.
  • Fund verification of worker sanitation at the farms that supply U.S. buyers: toilets, handwashing, worker health and water, checked by auditors who are not paid to pass.
  • Publish root cause analyses. FDA has promised an Outbreak Investigation Report. Taylor Farms should publish its own water results, from this year and from 2013.
  • Stop supporting FSMA rollbacks. Every “burden” removed this decade was a tool someone needed this summer.

Three petitions, and an invitation

There is one more thing I want from the industry: its help. I have drafted three citizen petitions to FDA, and I have not filed them yet. That is deliberate. Each one grows out of what went wrong this summer and in the outbreaks before it, each one would help the honest grower far more than it would hurt, and each one would carry more weight with FDA if the people who grow, pack and sell produce were on it with me. Here is a taste of what is in them, and why.

Let FDA sample at feedlots implicated in outbreaks. In 2018, FDA found the Yuma outbreak strain in an irrigation canal running past a feedlot of roughly 100,000 head, collected six samples inside the operation, and left. In Salinas in 2019 and 2020, the agency’s best evidence came from a cattle grate, a trough, a fence line and a roadside, places an investigator can stand without anyone’s permission. The first petition asks FDA to use its communicable disease authority under section 361 of the Public Health Service Act, the same authority behind the 1975 rule that still keeps small turtles out of pet stores, to require entry and sampling at an animal feeding operation once the agency has determined in writing that it is implicated in a multistate outbreak. It is deliberately narrow: no seizing animals, no regulating husbandry, no financial records, just a triggered, time-limited right to walk onto ground the agency has already identified as the probable source and take samples. Why should growers care? Because leafy greens farmers have been paying for their neighbors’ manure for twenty years, and right now nobody is allowed to look.

Name the firms implicated in outbreaks. On September 10, FDA closed a fourteen-person E. coli O157:H7 investigation, said for the first time that the food was romaine, and identified no grower, no processor, no brand and no state. Twenty-one months earlier it did the same in an eighty-nine-person romaine outbreak that killed one person. The agency has named farms and processors in these circumstances before; what it lacks is a written standard. The second petition asks FDA to establish by rule that the identity of a firm implicated in an outbreak is not confidential commercial information, to reach the whole supply chain from the field to the restaurant, and to name implicated firms whether or not there is a recall. Why should growers care? Because an unnamed outbreak is a category-wide outbreak. When the agency names one farm, it clears everyone else, and this summer California learned what it costs when nobody is cleared.

Tell people where recalled food went. Since 2008, USDA has published the retail consignee lists in Class I meat and poultry recalls. Congress told FDA in FSMA to consider doing the same. FDA wrote a guidance that does it sometimes. The third petition asks FDA to require it in every Class I food recall, including the restaurants, schools, hospitals and other institutions where bagged salad and shredded lettuce actually end up, and to classify every recall within ten days. FDA’s July recall notice for the Taylor Farms de Mexico lettuce listed customers only as CV, JB, MARK, MKTSD, PK, SUB, SY and TF. Why should the industry care? Because a recall nobody can act on keeps people sick longer and keeps the story in the news longer, and every day it stays there costs the whole category sales.

So here is the invitation. IFPA, Western Growers, the LGMAs, FMI, NGA and any grower, processor or retailer who wants in: email me and I will send you the drafts, tell me what is wrong with them, and help me fix it. If there is a better way to write the feedlot sampling trigger, a legitimate trade secret concern I have not accounted for, or a recall timeline that does not work in the real world, I want to hear it before FDA does. And then sign on. A petition from a plaintiffs’ lawyer is easy for an agency to set aside. A petition from a plaintiffs’ lawyer and the produce industry together is not. I will send the drafts to anyone who asks. My email is bmarler@marlerclark.com.

Twenty years

After spinach, the industry built something. It was imperfect and it was late, but it was real. This summer’s outbreak was bigger, the tools were weaker and the response was slower, and once again the growers who did everything right paid for the one who did not. FDA has sent a letter to industry asking for collaboration before the 2027 growing season. That is a short window.

In ten days, the produce industry will gather in Orlando. I would like to hear someone on that stage say three words: repeal the delay. Then I would like to hear them say “water.” And I would like to see their names next to mine on these three petitions.

About a week ago I posted drafts of two citizen petitions I intend to file with the FDA – one asking for a warning statement on cheese made from unpasteurized milk, and the other asking for a warning statement on raw and lightly cooked sprouts. I asked for comments and corrections. You delivered.

Many of you took the time to wade through more than 20,000 words of regulatory prose and tell me what I got wrong, what I missed, and what I could say better. Citations were fixed, facts were tightened, and arguments were sharpened (and a few were dropped). Some of you agreed with me, some of you did not, and both kinds of comments made the petitions better. Thank you.

The revised drafts, dated October 3, are here:

Because the comments keep coming – and because new outbreaks keep being announced while the drafts are being written – I am extending the time to weigh in through Thursday, October 15, 2026. I will file shortly after that.

Why these petitions are needed

The two petitions involve different foods, but they make the same basic argument and ask for the same modest thing: a sentence on the label. Neither asks FDA to ban anything. Adults who want raw milk cheese or raw sprouts would remain free to buy them. They would simply be told, at the moment they make the choice, what the government already knows.

  • The government already knows the risk. FDA has told children, the elderly, pregnant women and people with weakened immune systems not to eat raw sprouts since 1998, and has never withdrawn that advice. On raw milk cheese, FDA’s own science has found that the 60-day aging rule – written in the 1940s and the legal basis for shipping raw milk across state lines as cheese – does not reliably kill the pathogens it was presumed to control. FDA said in 2013 it would reevaluate that rule. It has not finished.
  • The advice has never been on the package. For nearly three decades the warnings have been delivered by press release, talk paper and web page. Almost no one reads those standing in a grocery aisle or at a sandwich counter.
  • The outbreaks have not stopped. FDA counts 52 sprout outbreaks and more than 2,700 illnesses between 1996 and 2020, and two more sprout outbreaks are open right now. On the cheese side, four outbreaks in 31 months – in cheese that complied with the 60-day rule – have sickened 43 people, hospitalized 17 and caused at least six cases of hemolytic uremic syndrome, many of them in children under five.
  • The label is often the only warning that arrives in time. In outbreak after outbreak, most people are sick before anyone announces a recall. Aged cheese sits in refrigerators for months. Cheese cut and wrapped at a deli counter and sprouts tucked into a restaurant sandwich carry no manufacturer’s label at all, which is why both petitions also ask for point-of-sale signs and online disclosures.
  • The precedent is exact. In 1998, after the Odwalla juice outbreak, FDA required a warning on unpasteurized juice. It later required a safe handling statement on shell eggs. Both sit in the same regulation – 21 CFR 101.17 – that these petitions ask FDA to amend. USDA has required safe handling instructions on raw meat and poultry since 1994, after Jack in the Box. The juice warning went from proposal to final rule in about 75 days.

A choice made without the material fact is not really a choice. These petitions are about putting that fact where the decision is made.

Please keep the comments coming

If you have not yet weighed in, please do. If you already have, look at the new drafts and tell me whether I got it right this time. I want to hear from scientists, regulators, public health folks, cheesemakers, sprout growers, retailers, lawyers, and especially from people who think I am wrong. Pointing out an error now is a lot more useful than pointing it out after the petitions are filed.

Call me or email me with your thoughts – bmarler@marlerclark.com or 1-800-884-9840 – any time before October 15.

Thanks again.

Bill

While I was on a plane heading back from Norway, on Friday evening Lena H. Sun of the Washington Post reported that FDA had closed its investigation of the largest Cyclospora outbreak in American history: the parasite was in wastewater leaving the Taylor Farms de Mexico plant and in a ditch at a grower that supplied it, and the agency could not say how it got onto the lettuce. Sun also reported two things the notice does not: FDA did not say when the samples were taken, and by the Post’s own analysis of FDA enforcement data, at least 3.1 million pounds of lettuce tied to the operation were recalled, against the previous record for a Cyclosporarecall, about 31,000 pounds of parsley in September 2025. A hundred times the old record, and it took a newspaper to add it up.

That has been the shape of this outbreak since July. The agencies published on a schedule. Sun published when she knew.

The ledger

July 10. Sun asked why nobody could say what food was spreading the parasite. I answered two days later that you cannot trace what you do not count: the official count was 843, and Michigan alone had more than 1,500.

July 14. Sun reported, from two people familiar with the investigation, that investigators were looking at Taco Bell and lettuce. CDC said that day it had found a likely link among cases in four states, and did not say where the people had eaten. The agencies said Taco Bell on July 16.

July 16. The night the agencies said Taco Bell and a single supplier of iceberg lettuce from Mexico, Sun and Rachel Roubein named the supplier: Taylor Farms. FDA’s page named the company on July 17. I wrote that day that the outbreak now had two names on it.

July 24. At five in the morning, Sun reported that CDC had tied four more states to the outbreak. CDC’s investigation page still said five, stamped July 16, and Reuters, picking up her story, could not reach CDC outside business hours. I asked that morning whether the Post was the new CDC. FDA’s advisory caught up later that day.

August 4. At 4:13 in the afternoon the Post named the six states that took the outbreak from nine to fifteen, from a fourth person familiar with the investigation. Every federal page said nine. FDA’s advisory, still titled for a nine-state outbreak, changed on August 5.

August 14. Roubein, Sun and Caitlin Gilbert reported, from four people with knowledge of the matter, that FDA investigators had reached the Taylor Farms plant that week, a month after the agency named the company. The interval was the headline, I borrowed it, and the explanation came from Donald Prater, FDA’s acting deputy commissioner for food, in an interview with the Post.

September 11. Roubein reported, from two people familiar with the matter, that federal officials were about to declare the outbreak over. They did later that day. I learned it from the Post.

What the attributions tell you

Count them. Two individuals familiar with the investigation. A fourth person familiar. Four people with knowledge of the matter. Two people familiar with the matter. They are people inside the investigation telling a reporter what their own agencies’ pages would not say for another day, or another twelve. The information existed every time; it reached the public through a paywall because the federal pages were waiting on a schedule. The reporting is not the problem. Sun and Roubein have been right every time I could check, which is every time.

Congress reads her too. Senator Warren’s August 13 letter to Secretary Kennedy footnotes two of Sun’s stories. Fifteen letters from members of Congress have gone to FDA, CDC, HHS, USDA and Taylor Farms since July. Not one has a published answer, and no committee has noticed a hearing.

The next story is on Capitol Hill

FDA closed its file on Friday and put ten commitments in writing for the 2027 growing season. That is the agencies’ chapter. Congress has not written its own, and the next central Mexico growing season begins this fall. Sun’s next story is in the Agriculture-FDA appropriations subcommittees, the HELP Committee and the Permanent Subcommittee on Investigations. Here are the problems and the fixes, as I set them out Saturday on Food Safety News, each with a question.

The traceability rule. The administration pushed compliance from January 2026 to July 2028, and in November 2025 Congress barred FDA from spending a dollar to enforce it before then. Michigan asked Taco Bell three times over six days for a supplier list the rule would have required in a day; FDA had complete records on July 13, and the lettuce shipped through July 16. The fix is to repeal the rider, as eleven consumer and food safety organizations asked in August. The delay was requested by retail, grocery, restaurant and distribution associations, whose own lobbying filings say so. The question: the Senate Appropriations Committee’s report said FDA could not delay past July 2028; the enacted bill says FDA cannot enforce before it. Who put Section 780 in, and why has no senator on that subcommittee asked?

Counting. On July 1, 2025, CDC cut FoodNet from eight pathogens to two, and Cyclospora was dropped. CDC counted 19,883 confirmed cases this season, roughly 7,000 of them never tied to any food. The nominee to run CDC, asked in July whether she would restore the reporting, would not commit. The fix: FoodNet back to eight pathogens, one national number, the illness reportable in every state, and the parasite on the national wastewater panel. The question: what did dropping six pathogens save, and who signed it?

The states. In March 2025 HHS clawed back $11.4 billion from state and local health departments, $8.9 billion of it epidemiology and laboratory money. A court restored it only for the twenty-three states and the District that sued. Ohio, Indiana and West Virginia, three of the five original outbreak states, were not among them. The fix is to fund the grants for all fifty; the question is how Ohio counted its share of this outbreak on what was left.

The science. Congress declined to fund two of the three Cyclospora projects at USDA’s Beltsville research center for fiscal 2026, and the parasitologists there were told to relocate or lose their jobs. Kalmia Kniel at the University of Delaware asked USDA four years running for about $10 million to build a water test for this parasite and was turned down. USDA ended the advisory committee that wrote the government’s 2023 report on this parasite, saving $225,000. The fix: fund Beltsville and the Delaware test, rehire the parasitologists, bring the committee back.

The testing program. USDA’s Microbiological Data Program tested about 15,000 produce samples a year for $4.5 million until it was zeroed out in 2012 at the produce industry’s urging. USDA’s own economists price this parasite at $5.3 million a year. The fix is to restore it with the parasite on the panel; it costs less than the disease.

Inspections. Congress set a goal of 19,200 foreign food inspections a year by 2016. In 2025 FDA made about 1,100. Before this summer it had not visited the Guanajuato plant since 2019, and it laid off the staff who plan foreign inspection travel before reinstating them in August after the New York Times asked. The fix: inspect foreign plants on a schedule and fund the inspectors. Susan Mayne, who directed FDA’s food safety center, gave Sun three more on Friday: research on treating agricultural water, a produce safety expert stationed in Mexico, and testing of lettuce before it crosses the border next summer.

Transparency. A petition from STOP Foodborne Illness asking FDA to name every firm in an outbreak has sat at the agency since February; the growers in this one are still unnamed. The genotype comparison between the water and the patients is pending, and Representative Garcia’s July 27 request for Taylor Farms’ records back to 2013 has no public answer. The fix: grant the petition, publish the investigation report and the genotype comparison with a date, and name Cyclospora in the agricultural water rule, whose only numeric standard is for E. coli, a test that cannot find a parasite.

The hearing. The Senate’s Permanent Subcommittee on Investigations held a hearing on this parasite in 1998, and its chairman can issue a subpoena. The Secretary said the outbreak was under control on July 21, when the count was 1,644. For the fifteen offices that wrote, the question is what they have gotten back. For the committees with gavels, it is why the only hearing room this outbreak reached was a confirmation hearing.

Sun has spent a summer telling the public what the government knew before it said so. The sequel is what Congress does about it, and it has a calendar. I asked the beef industry to put me out of business in 2002, and it largely did; I have been asking the lettuce business for nearly thirty years. Ask the appropriators, ask the members who signed the letters, ask the ones who did not, and print the answers next to the dates. The next crop goes in the ground this fall.

Saturday, October 3, 2026

This is the thirteenth Saturday column. Last Saturday’s edition ran under the headline A Raw Milk Cheese Labeled as Aged the Sixty Days the Rule Requires Made Thirteen People Sick; its point was that the recalls were arriving after the outbreaks were already over. This week the government closed the books. On Friday FDA ended its investigation of the largest Cyclospora outbreak ever recorded in this country, 12,883 people sick, and said for the first time where it found the parasite: in a tank of wastewater leaving the Taylor Farms de Mexico plant, and in a drainage ditch at a grower that supplied it. The same afternoon it ended the jalapeño Salmonella outbreak at 488 sick, with a grower traced to a Mexican state and no route of contamination traced at all. The day before, CDC closed the hepatitis A outbreak in frozen blood clams at 37 sick and 30 hospitalized, with the importer never named. Three files, 13,408 people, and not one closing page that says how the pathogen got into the food. The files that stayed open did the opposite of closing. The joint FDA–USDA E. coli investigation went from 25 sick to 38 with no food named, a new Listeria file opened at 11, and CDC’s alfalfa sprout count went from 55 to 76 while FDA’s page for the same outbreak still says 55. Closed without a cause, open without a food.

FDA closed the iceberg lettuce investigation on Friday and, for the first time, said where it found the parasite. The closure page carries the final count, 12,883 confirmed ill, 570 hospitalized and two dead in 21 states, with the last illness beginning August 17, and two sentences the agency had not written before: one positive environmental sample came from a tank that held outgoing wastewater from the Taylor Farms de Mexico processing facility, and the second from a drainage ditch at an iceberg lettuce grower identified in traceback. No lettuce ever tested positive. Genotyping of both samples is complete, and the page does not say whether they match the people. FDA’s own conclusion is that there is not enough evidence to conclusively determine how contamination occurred. Taylor Farms told ABC News the grower is no longer in its supplier program and that the Mexico facility remains closed. I wrote Friday about what FDA found, wrote again after the New York Times put the finding in plain words about what the company has said and has not, and in Food Safety News this morning about what has to happen before the 2027 lettuce is planted. The one thing this week I will credit without qualification is that FDA also put ten commitments in writing for the 2027 growing season, from an outbreak investigation report to laboratory capacity in Mexico to standardized genotyping with CDC, and the Office of Produce Safety deserves the credit. Four other Cyclospora files on FDA’s table are still marked active with their outbreaks ended, 313 people and no food among them, and the season total CDC stopped updating weekly stands at 19,883. Fifteen congressional letters still have no published answer.

The jalapeño outbreak closed the same afternoon, with a grower traced to a state and a cause traced to nobody. FDA’s final page counts 488 sick in 34 states, 65 hospitalized and no deaths, onsets June 16 to August 26. Of 308 people interviewed, 281 ate at a Mexican-style restaurant, Chipotle and QDOBA among them. Traceback found a common grower in Sinaloa that supplies Coast Citrus Distributors; by the time inspectors arrived no jalapeños were left to sample, the environmental samples were negative, and the page says in one sentence that FDA was unable to determine a route or source of contamination. My post on the Salmonella Javiana close is up, and Phyllis Entis wrote up both closures side by side. CDC closed the week’s third file on Thursday: the hepatitis A outbreak in La Serranita frozen blood clams from Ecuador ended at 37 sick, 30 of them hospitalized, in four states, with the New Jersey importer and the New York dealer never named on any federal page. I wrote Monday about whether those clams were illegally imported; New Jersey says yes, and federal law makes it a harder question.

The Graziers raw milk cheese recall reached FDA’s recall page on Tuesday, five days after the company announced it, and it still has no lot codes. The notice lists eleven products, from eight-ounce squares of Monterey jack to 40-pound cheddar blocks, distributed nationwide, and says the investigation has not been narrowed to a specific lot, date code or production date; everything is recalled. The CDC and FDAcounts have not moved since September 25: 13 sick in nine states, eight hospitalized, three with hemolytic uremic syndrome, more than half of them five or younger, onsets July 7 to August 26. Phyllis Entis had it first as the second multistate raw milk cheese outbreak this year and the third since 2024. I wrote Thursday about why thirteen is a floor, why the HUS rate is so high, and the Listeria recall at the same plant thirteen months earlier. Marler Clark represents the family of a child who developed HUS after eating Graziers cheese, and families from last fall’s Twin Sisters Creamery outbreak. The draft citizen petition for a warning statement on raw milk cheese and completion of FDA’s 60-day-rule reevaluation is open for comment through Friday, October 9, and the twenty-five-year outbreak record behind it ran Monday. New York added a footnote Wednesday: Agriculture and Markets found Listeria in raw milk from Big Brook Farm in Oneida County, the farm’s second positive since 2024, and barred sales.

CDC’s alfalfa sprout count went from 55 to 76 on Thursday; FDA’s advisory for the same outbreak still says 55. CDC’s October 1 update adds 21 sick, two hospitalizations and a sixteenth state, Oregon. Onsets now run from May 31 to September 7, sixteen days after the August 22 recall; 64 people have E. coli, nine have Salmonella Agona and three have both, and CDC writes that recalled seed lot SAL2-24JB may be contaminated with multiple pathogens. FDA’s advisory is stamped September 24 and carries 55 sick in fifteen states, and its table row has not moved either. I put the seventy-six in context Thursday: 72 of them were already sick before the public was told anything. The seed kept finding new doors: Todd’s Seeds recalled again, 37,481 units of sprouting seed and mixes containing the lot, a recall that appears in FDA’s enforcement database and nowhere on its recall page. In the other sprout outbreak, EAK Distribution recalled HOME GROWN broccoli sprouting seed sold on Amazon nationwide after FDA found Salmonella in the seed; the same bulk lot went to Evergreen Fresh Sprouts, the grower in the 32-person broccoli sprout outbreak, which has not moved since September 24, Washington still at 24. That is 108 sick across the two sprout outbreaks, and the second time in six weeks that contaminated seed reached home sprouters, which is the argument for the sprout warning label petition, also open for comment through October 9. The twenty-five-year sprout record ran Tuesday.

Eleven open files on FDA’s outbreak table name no food, and together they carry 675 people.The table was re-stamped September 30. The joint FDA–USDA E. coli O157:H7 file, FDA #1421 and FSIS 2026-04, went from 25 sick to 38 in a week and FDA began collecting samples; the product line still reads Not Yet Identified. A new Listeria file, #1415, opened at 11 with traceback only, #1417 went from 40 to 44, and #1379 from 10 to 11, which puts 66 people in three files with Listeria and no food named. On Tuesday I wrote about two calls in twenty-four hours, both culture-confirmed E. coli, both told they are part of a multistate outbreak, both buyers of a meat-and-vegetable sandwich wrap from the same chain in different states; a wrap would explain why one outbreak carries two reference numbers. The firm’s epidemiologists have the records requests out, and I said what the count looked like the next morning and asked the agencies for states, onsets and whether a mixed product is under investigation. CDC’s dashboard, re-dated the same day, lists 33 active multistate investigations, 18 of them Salmonella, down from 38 the week before.

The frozen blueberry recall expanded a third time on Friday, eleven days after CDC declared that outbreak over. Frutas y Hortalizas del Sur of San Carlos, Chile, added Great Value Organic Blueberries and Organic Triple Berry Blend at Walmart in 30 states and Puerto Rico, and Trader Joe’s Organic Mixed Berry Blend in 26 states and the District of Columbia, with best-by dates that run to June 2028. CDC closed the E. coli O145:H28 outbreak September 21 at 17 sick and six hospitalized in Florida and Georgia, no deaths. A suit filed in Pinellas County Circuit Court alleges a Largo woman died June 20 after eating GreenWise blueberries bought at Publix, thirteen days before the first recall; nobody has said whether she is among the seventeen. My post on the expansion is on Food Poison Journal.

Phyllis Entis at eFoodAlert kept the ledger, and the items that matter this week are mostly about what the notices leave out. Gias Foods expanded its bettergoods Lemon Alfredo Fettuccine recall at Walmart to every lot on Thursday, Listeria found by Washington and Florida state sampling, best-by dates into September 2027, no illnesses; the two-lot recall was September 15. New York’s Agriculture and Markets announced a nationwide Listeria recall of Starfish Seafood Cold Smoked Seafood Trio from a Brooklyn distributor that appears on no FDA page (October 1). Coco G’s Snack Foods popcorn, recalled in May for Salmonella, was classified Class I in FDA’s enforcement report on September 29 and appears nowhere else (September 30). G.L. Mezzetta recalled one lot of peperoncini for a pest contaminant ten days after a consumer posted video of what appeared to be a mouse in the jar; the notice never says mouse, and WinCo pulled salsa bean dip trays in eight states that used it. I wrote about the notice that never says so and about thirteen recalls since 2009 where the contaminant was an animal. On the meat side, FSIS recalled 189 pounds of Colombian pork-crackling chips on Friday (October 2), and Star Meat Delivery denied using a fake USDA mark on 167,639 pounds of uninspected meat, saying establishment number 1363 was legitimately reserved for it during the application process; USDA answered that the company never completed the process, that the product surfaced at a South Carolina retailer, and that a retail list would be added to the notice. No FSIS public health alert was issued in September. FDA posted a warning letter to Middlefield Original Cheese Cooperative in Ohio on September 29, Listeria in 14 of 80 environmental swabs after two 2025 recalls; the letter is dated March 6, 207 days earlier.

The nominee to run FDA will not get a vote before the midterms, and the recall count became a campaign line. STAT reported that the Senate HELP Committee left town without scheduling a vote on Dr. Heidi Overton and does not return until November 9; she sat for her hearing September 24 and could not name the office that investigates outbreaks. FDA’s own account on X said the same week that 1,836 recalls in fiscal 2026 is among the lowest totals in a decade and warned readers off the fake news; CBS News and KFF Health News ran the numbers, and I explained Monday why recall counts are a lousy yardstick. Susan Mayne, who ran FDA’s food center for eight years, asked why FDA’s recruiting drive for thirty advisory committees includes no committee on food; the microbiological criteria committee was eliminated last year. House Democrats put their health care agenda on paper with sixteen food bills, one of them about a pathogen. A federal judge in San Francisco extended the pause on USDA’s relocation of about 2,600 employees, Beltsville’s Cyclospora researchers among them, to October 13. Annie Waldman at ProPublica told the story of two little girls, an unlicensed goat dairy and a cease-and-desist letter that let the milk keep selling as pet food; hats off, and the firm has filed suit for both families. And the firm is still hiring an associate and a paralegal.

That’s the week — three outbreaks closed with 13,408 people in them and not one page that says how the pathogen got into the food, eleven files open with 675 people and no food named, 108 sick across two sprout outbreaks with the seed still turning up in new bags, and a raw milk cheese that put three small children into kidney failure recalled nationwide with no lot codes. Two draft petitions are open for comment through Friday, October 9; read them and tell me what I got wrong. Check back next Saturday.

About Bill Marler

William “Bill” Marler has spent more than thirty years as a food safety lawyer and advocate—work that began with the 1993 Jack in the Box E. coli outbreak and has never really stopped since. In the years that followed, he has represented victims in nearly every major foodborne illness outbreak in the United States. That case, and the movement it launched, is the subject of the book “Poisoned” and the Emmy Award–winning Netflix documentary of the same name. Bill’s work has been profiled in The New Yorker (“A Bug in the System”), the Seattle Times (“30 years after the deadly E. coli outbreak, a Seattle attorney still fights for food safety”), the Washington Post (“He helped make burgers safer. Now he’s fighting food poisoning again”), and many others.

Dozens of times a year, Bill speaks to industry, regulators, and universities across the United States, Canada, Europe, Africa, China, and Australia about a simple idea: outbreaks are preventable. He has testified before Congress on the Food Safety Modernization Act and teaches food safety at institutions including the Harvard T.H. Chan School of Public Health. He writes regularly about food litigation and food safety at Marler Blog, and in 2009 he founded Food Safety News, which he continues to publish.

On Friday, Christina Jewett, Alice Callahan and Caroline Hopkins Legaspi of the New York Times reported FDA’s finding in the terms it deserved. Inspectors found Cyclospora in water from a tank of outgoing wastewater at the Taylor Farms de Mexico plant that chopped the lettuce, and in sediment from a drainage ditch at an iceberg grower that supplied it, and the Times called that strong new evidence that the summer’s outbreak came from the company’s lettuce. That is the right reading. Jewett, who covers FDA, and Julie Creswell wrote the Times story in August, with Legaspi contributing, that put dates on Michigan’s three requests to Taco Bell for a supplier list and on the rule that would have required an answer in a day. When I think the Times is behind the record, I say so. On Friday it was not, and the three of them deserve the credit.

The Times has the numbers right: nearly 13,000 illnesses, 570 hospitalizations and two deaths in twenty-one states, the largest cyclosporiasis outbreak on record. FDA’s final count is 12,883. CDC declared the outbreak over on September 11, and FDA closed its investigation on October 2. I wrote up FDA’s update and its ten commitments for 2027 on Friday and will not repeat them.

What FDA confirmed

FDA’s words, nearly: the positive samples show the parasite was present where the lettuce was grown and processed, and they add to the strong epidemiological data and traceback convergence supporting the link between this outbreak and the recalled shredded iceberg lettuce from Taylor Farms de Mexico. In July, Michigan’s interviews and Taco Bell’s supply chain records pointed at one supplier. Taylor Farms recalled the lettuce itself, 236,192 cases sold to Walmart, Subway, Jack in the Box, Taco Bell and Sysco. On August 20, CDC said genotyping showed the parasites from sick people were genetically related to one another. Now the parasite is in the water leaving the plant and the ditch beside the field. The source question is closed. The epidemiology, the traceback, the company’s own recall, the patient genotyping and the environmental samples point to the same place.

What FDA says it cannot yet determine is how the parasite got onto the lettuce, and whether the genotypes of the two environmental samples match the patients. Both have been typed; the comparison is still being analyzed. Those are mechanism questions, not the source question, and the company’s statement runs the two together. The Times drew one more inference: Cyclospora cayetanensis spreads only through human feces, so the parasite in that tank and that ditch came from an infected person on or near those operations.

What Taylor Farms told the Times

The company told the Times that the plant has not reopened since production was suspended in July, that it is working with its research team on enhanced safety protocols from the field to the facility, that the independent farm where one sample was found is no longer in its supplier program, and that trust is earned by what it does. It also emphasized, as the Times put it, FDA’s conclusion that there is not enough evidence to determine where the contamination occurred. It has said before that it spends more than $200 million a year on food safety.

Read the one word. FDA said how. The company said where. Where has an answer: the parasite was found at a grower in its supplier program and in a tank of its own wastewater. Changing how to where turns an open question about mechanism into a closed question about geography, and the company then answers the geography by dropping the grower. The tank is at the plant. The plant is the company’s.

What the company has said since July, and what it has not

The record is short. On July 17 the company said FDA’s tracing pointed to a specific independent farm in Mexico. FDA now says several farms and the plant, and found the parasite at both. In August the company told the Times it runs a fecal indicator testing program covering all of its water sources and that every result was negative. Indicator tests look for bacteria; FDA’s own method for Cyclospora in agricultural water looks for the parasite. The company’s information hub, last modified September 1, says roughly 2,000 of its own samples found no positives, that its water is tested for indicator organisms, and that decisions will wait for its third-party review and the FDA investigation. The FDA investigation has concluded. The page has not moved, and the newsroom lists no release on the outbreak.

Here is what the company has not said, to the Times or to anyone. How the parasite got into a tank of water leaving its own plant. Whether it ever tested that wastewater for the parasite rather than for indicator bacteria. What the independent third-party review it promised in July found, and whether it will be published. What enhanced safety protocols means in practice: a treatment step, a different water source, a worker health program, or all three. When the plant will reopen and on what showing. And an answer to Representative Garcia’s July 27 request for its records back to 2013.

In 2013, after the last Cyclospora outbreak traced to this plant, the company said it would test its water, and FDA reopened the plant on that promise. Thirteen years later I could not find one public result. Enhanced safety protocols is also a sentence. The question for 2027 is whether it stays one.

What I told the Times, and what I have been asking for since July

The Times quoted me on five things. First, there is more to learn from the environmental samples and from patients’ stool, and FDA should publish the genotype comparison with a date attached. Second, before the next growing season the federal government should pay to develop a routine water test for Cyclospora, which does not exist. Kalmia Kniel of the University of Delaware and her colleagues asked USDA for about $10 million to build one four years running and were not funded, as the Times reported in August; USDA says its reviewers chose other projects. Third, enforce the rule requiring supply chain records within twenty-four hours of a product being implicated. Michigan asked Taco Bell on July 1, July 3 and July 6, and FDA had complete records on July 13. The traceability rule that would have required an answer in a day was delayed to July 2028, and Congress then barred FDA from spending a dollar to enforce it before then. The Times itself found that the delay may have slowed the response. Fourth, inspect. FDA had not visited the Doctor Mora plant since 2019; FSMA set a goal of 19,200 foreign inspections a year by 2016, and in 2025 the agency made about 1,100.

Fifth, an answer for the roughly 7,000 confirmed cases this summer that were never tied to any food. CDC counted 19,883 laboratory-confirmed domestic cases with onsets between May 1 and August 31. Of those, 12,883 are in this outbreak. The rest have no name.

None of that is new. Since July I have asked FDA to grant STOP Foodborne Illness’s pending petition and name the growers, asked Congress to unwind the enforcement bar as eleven consumer and food safety organizations did in August, asked for the shelved Beltsville Cyclospora research to be funded, and asked for a hearing. Fifteen congressional letters have gone to FDA, CDC, HHS, USDA and Taylor Farms since July. None has a published answer, and no committee has noticed a hearing. Jim Gorny, who retired from FDA in 2024, told the Times that without knowing how the lettuce was contaminated, the new measures are the same strategies tried again in hope of a different result; Frank Yiannas, who ran food policy at FDA from 2018 to 2023, hopes this summer becomes a catalyst. I would put it less gently. The parasite was found in the water. The company has a page that says it is waiting for FDA. FDA has a list of ten commitments and one date. Congress has a rider.

Marler Clark represents about 500 people sickened in this outbreak and has filed suits in Ohio, Michigan, Kentucky, Kansas, Missouri and Illinois.

Three reporters put FDA’s finding, the company’s answer and the open questions in one place on Friday, which is more than the company’s website has done since September 1 and more than any committee of Congress has done since July. The next thing worth printing is the genotype comparison, and after that the investigation report that says what the water looked like. Jewett, Callahan and Legaspi will read both. The company should assume so.

Three questions keep arriving in my inbox: what did industry say, did Congress ever get an answer, and what are the final numbers. Here are the short answers, then the record behind each. Taylor Farms gave a reporter a statement and changed nothing on its own website. The largest produce association answered FDA’s letter to industry with a letter of its own asking FDA to do the root cause work. Fifteen letters from members of Congress have produced no published answer and no hearing. The federal numbers are final. The state numbers are not, and nobody has said they are.

What Taylor Farms said, and where it said it

When FDA closed its investigation on October 2 and reported the parasite in a tank of outgoing wastewater at the Taylor Farms de Mexico plant and in a drainage ditch at a traceback grower, a Taylor Fresh Foods spokesperson told ABC News that the central Mexico facility remains closed while the company works with its research team on enhanced safety protocols, that the independent grower where a positive sample was found is no longer part of its supplier program, and that, as FDA noted, there is not enough evidence to determine how contamination occurred. The company said it remains concerned for the people and families affected. I wrote up what FDA found here.

Read it for what it does not address. The grower with the positive ditch is out. The company’s own plant, where the positive wastewater tank sits, gets the phrase enhanced safety protocols and nothing more specific. The plant has been closed since July 18.

Then go to the company’s Cyclospora information hub. As of this morning it carries a September 1 modification date. It says twice that according to FDA there are no confirmed positive sample results for product testing as of August 27, 2026. It says that confirming a specific product requires a positive laboratory test of the product, that the company’s own testing of roughly 2,000 samples in central Mexico since May returned zero positives, and that water sources are tested for indicator organisms. It says decisions about the future will be based on what the independent third-party review and the FDA investigation find. The FDA investigation has now found, and the page has not moved. The company’s newsroom lists no release on the outbreak at all; the most recent entry is a June 30 scholarship announcement.

In fairness, FDA’s two positives were environmental samples, not product, so the product-testing sentence is still technically true. It is no longer the point; the spokesperson did not repeat it on October 2. And the line that water sources are tested for indicator organisms now sits next to a federal finding of the parasite itself in water leaving the plant.

What the trade associations said

FDA’s letter to industry, dated September 11 and posted September 16, asked every link in the supply chain to develop new science-based practices against this parasite. On September 16, the International Fresh Produce Association sent its own letter to Acting Commissioner Diamantas asking for clearer communication, a transparent root cause analysis, and FDA support for research on traceability in the produce industry. Two letters crossed in the mail, each asking the other side to do the work. IFPA’s Cathy Burns said separately that the association supports putting the Food Traceability Rule into effect without further delay, while wanting workable standards, and IFPA had earlier proposed an FDA-led public-private task force to run the root cause analysis once the investigation closed. Credit where it is owed: that traceability position is the right one, and it puts the growers on the opposite side from the retail associations whose lobbying filings asked Congress for the delay.

IFPA closed its own Cyclospora page on September 14, which I wrote about at the time. As of this morning I can find nothing from IFPA, Western Growers, the leafy greens marketing agreements or Taco Bell on FDA’s October 2 close. Give it until Monday.

Fifteen letters, zero answers, one hearing room

On September 5 I listed thirteen letters from members of the House and Senate, eight of them with deadlines that had already passed. Representative Dingell wrote to USDA on September 14 about the shelved Cyclosporaresearch at Beltsville, and Representative DeLauro wrote to Secretary Rollins on September 17 with twenty-four colleagues from both chambers on the same subject. That is fifteen. I updated the list on September 16 and checked again this week. I cannot find a published answer to any of the fifteen. Not to Representative Garcia’s request for facility records going back to 2013, not to Senator Blumenthal’s two July 30 letters as ranking member of the Permanent Subcommittee on Investigations, not to Senator Warren’s nine questions due August 27.

The official answers that exist were given to reporters. HHS’s press secretary told The Hill on August 13 that the response was guided by science and that outbreak investigators were untouched by the staffing cuts, which I recorded then. A USDA spokesperson told Fox stations in late August that Congress did not fund two of the three Beltsville Cyclospora projects for fiscal 2026 and that the third is moving to Iowa. FDA told the New York Times that laid-off travel and foreign-coordination staff had been reinstated. FDA’s October 2 update is the first substantive public answer to any records question Congress asked, and it was addressed to nobody.

No committee has noticed a hearing. The one hearing room this outbreak entered was the Senate HELP Committee’s September 24 confirmation hearing for Dr. Heidi Overton, where it came up in questions and, as Food Safety Magazine reported, the nominee could not identify the FDA office that runs outbreak investigations. The Permanent Subcommittee on Investigations held a hearing on this parasite in imported raspberries in July 1998. Its chairman can issue a subpoena. Its ranking member has already written twice.

The count: the federal numbers are final, the state numbers are not

The outbreak FDA and CDC attribute to the recalled lettuce closed at 12,883 illnesses in 21 states, 570 hospitalizations and two deaths, with the last illness beginning August 17. CDC declared it over on September 11. FDA closed its investigation October 2.

The season count is the number to remember. CDC’s case data page reported 19,883 laboratory-confirmed cases acquired in the United States with onsets between May 1 and August 31, 1,064 hospitalizations, two deaths and 49 states plus the District of Columbia as of its September 15 update, against 1,180 confirmed cases for the same stretch of 2025. Another 9,765 confirmed cases were still being sorted into domestic or travel, and CDC said it was aware of 6,594 more. That is 36,242 people the agency knows about. CDC stopped weekly updates after September 15.

Subtract the outbreak from the season and 7,000 confirmed cases have no food named. FDA’s outbreak table still carries four Cyclospora files, 313 people in all, where the outbreak has ended and no food was ever identified.

The states are a different series, and they are not done. The last time I added up all fifty, on September 3, forty-one states published a 2026 count and those summed to 33,457; Food Safety News carried the tally at 33,548 on September 12. Michigan alone reported 14,718 cases and 366 hospitalizations as of September 3, and its page now says it updates on the first Thursday of each month, which makes October 1 the first new number since the season closed. Indiana closed its count at 2,104 pending a formal close. Nearly every other state stopped updating in late August or September, and none has posted a number it calls final. I will publish a fifty-state closing table, each state’s last posted count and date beside CDC’s, as soon as Michigan’s October figure is in hand.

Marler Clark represents hundreds of people sickened in this outbreak and has filed suits in Ohio, Michigan, Kentucky, Kansas, Missouri and Illinois.

Everything that could be closed has been closed. The season closed August 31. CDC closed the outbreak September 11. FDA closed the investigation October 2. The company’s page closed on August 27 and has not reopened. The fifteen letters never opened at all. Two people are dead, 570 went to a hospital in the one outbreak that has a name, 7,000 confirmed cases have none, and the nearest thing to an answer under oath is a confirmation hearing where the parasite was a question. One subpoena would change that.

On October 2, 2026, FDA closed its investigation of the Cyclospora outbreak linked to shredded iceberg lettuce from Taylor Farms de Mexico. The final count is 12,883 illnesses in 21 states, 570 hospitalizations and two deaths, with the last illness beginning on August 17. CDC declared the outbreak over on September 11. Today’s update is the first time FDA has said what its inspectors found in Mexico, and it is the first laboratory finding from the growing and processing environment in this outbreak.

FDA’s traceback identified several farms and one processing facility in Mexico that supplied the recalled lettuce. Investigators inspected and sampled multiple growers and the plant in coordination with Mexican officials, collecting water, soil, sediment and lettuce. Two samples tested positive for Cyclospora. One came from a tank holding outgoing wastewater at the Taylor Farms de Mexico processing facility. The other came from a drainage ditch at an iceberg lettuce grower identified in the traceback. No lettuce sample tested positive. The only product positive ever announced, a border sample on July 18, was withdrawn as a false positive the next day.

FDA also said what the two positives do and do not show, and it deserves credit for saying it plainly. They show the parasite was present where the lettuce was grown and processed, and they add to the epidemiology and the traceback that pointed at this lettuce in July. They do not show how the contamination happened. FDA says there is not enough evidence to determine that. Genotyping of both samples is complete, and FDA is still working out how those results relate to the genotyping of specimens from sick people.

The last time FDA sampled this operation

In 2013, after a Cyclospora outbreak that sickened 631 people in 25 states, FDA and five Mexican agencies conducted an environmental assessment at Taylor Farms de Mexico in Guanajuato. By my reading of that assessment, about 835 samples of lettuce, water and worker stool were tested and none was positive. The one sample that could not be read was recycled wash water from inside the plant, too turbid for the filtration method. FDA could not determine how or where the parasite got into the product. Taylor Farms said it would test its water. Thirteen years later I could not find one public result. I have written about the two federal documents and about this company’s outbreak and recall record since 2009.

Put the two investigations side by side. In 2013, the one sample the method could not read was water inside the plant. In 2026, one of the two positives was water leaving the plant, and the other was water beside a field. FDA moved its sampling upstream this time, from the plant to the growers, which it announced on August 20, and a field is where a waterborne parasite gets onto lettuce. A positive in a wastewater tank is water that has already washed lettuce. It does not tell you whether the parasite arrived on the leaves or in the building. It tells you it went through the building.

Two statements from this summer belong next to that. On July 17, Taylor Farms said FDA’s tracing indicated that a specific independent farm in Mexico was responsible. FDA now says its traceback identified several farms and the processing facility, and the positives came from a grower and from the plant’s own wastewater. In August the company told the New York Times that it runs a fecal indicator testing program covering all of its water sources and that every test had been negative. Indicator tests look for bacteria, not for this parasite. FDA’s own method for Cyclospora in agricultural water looks for the parasite itself, and a sample reported positive for Cyclospora can only come from a test that does. Negative indicator results on incoming water and a positive parasite result on outgoing water are not in conflict. They are two different questions, and only one of them is the right question.

The sentence everyone is waiting for

On August 20, CDC said genotyping showed the parasites from sick people were genetically related to one another, which closed off the argument that this summer was unrelated cyclosporiasis piling up at once. It did not tie anyone to the lettuce. Today FDA says the two environmental samples have been genotyped and it is still analyzing how they relate to the clinical results. That is the comparison that matters. If the parasite in the wastewater tank and the ditch matches the parasite in the patients, the loop that 2013 never closed is closed. If it does not match, that is worth knowing too. FDA should say which, and say when it will say.

What FDA says comes next

On September 11, the day CDC closed the outbreak, FDA posted a page titled Post-Outbreak Response Activities. It lists a meeting with Mexican officials the week of September 8 about a prevention framework, a continuing root-cause investigation, a letter to industry, and a promise to fold lessons learned into the Cyclospora Prevention, Response and Research Action Plan in the coming months. Today’s update says that page will also carry prevention activities planned before the 2027 growing season. The letter, signed September 11 by Donald Prater, the acting deputy commissioner for food, reminds growers, packers, importers and retailers of rules already on the books and asks every link in the supply chain to develop new science-based practices against this parasite. The Action Plan it points to was written in 2021, and its page was last updated on June 30, 2023. It opens by describing roughly 3,000 domestically acquired cases over three years. CDC confirmed 19,883 this season as of its September 15 update.

Of those 19,883, 12,883 are in this outbreak. Roughly 7,000 confirmed cases have no food named, and FDA’s own outbreak table still carries four Cyclospora files, 313 people in all, where the outbreak has ended and no food was ever identified. The National Advisory Committee on Microbiological Criteria for Foods, which FDA charged in 2021 with the research questions the letter now hands to industry, delivered its Cyclospora report in 2023 and was terminated in 2025.

Before the first head of lettuce goes in the ground for 2027

Say whether the genotypes match. The environmental results exist. Compare them to the patients, publish the answer, and put a date on it.

Publish the root-cause findings. Which farms, how the water was used and treated, what the soil and sediment showed, and what was found at the plant. FDA published the 2013 assessment. It should publish this one.

Test water for the parasite, not for a stand-in. FDA has a method for Cyclospora in agricultural water. The pre-harvest water rule names no parasite. Name it, and say what a grower is supposed to do with a positive.

Records in twenty-four hours. Michigan asked Taco Bell for its supplier list on July 1, July 3 and July 6, and FDA had complete records on July 13. The traceability rule that would have required an answer in a day was delayed to 2028, and Congress then barred FDA from spending a dollar to enforce it before then. FDA can ask Congress to unwind that, and the part of industry that is already compliant can ask with it.

Inspect on a schedule, not after an outbreak. Before this summer, as the Times reported, FDA had last visited the Doctor Mora plant in 2019. A grower that ships to thirty-one states should expect a visit more often than once in seven years.

Say what the 7,000 are. If the clusters cannot be solved, say that, and say what would have been needed to solve them.

Marler Clark represents hundreds of people sickened in this outbreak and has filed suits in Ohio, Michigan, Kentucky and Illinois.

FDA found the parasite in the water this time. In 2013 it could not read the water, and nothing public ever came of the promise to test it. The difference between those two years is a sample result and a page that says what it means. The difference between this year and next will be whether anybody is required to do anything about it before the 2027 lettuce is planted.