What FDA actually wrote about Taylor Farms de Mexico in 2013, what FDA wrote about Taylor Farms Colorado in 2024, and why both were opened only after people were already sick.
And, although I looked, I could not find other FDA inspection reports – I assume they exist?
Taylor Farms has told the public this week that FDA “has not identified a single positive product test result for Cyclospora.” That is true. It is also the same thing the company and the agency were able to say in 2013, and again in 2024. Two federal documents explain why that sentence has never meant what the company wants it to mean.
Where this Cyclospora Outbreak Stands
As of today, CDC counts 1,644 people infected with Cyclospora who reported eating at Taco Bell across five states — Indiana, Kentucky, Michigan, Ohio and West Virginia. Illness onsets run from May 13 through July 13, 2026, with 94 hospitalizations and no deaths. That is the confirmed cluster, and it is the smallest number in this story. Nationally, CDC reports more than 1,645 laboratory-confirmed domestically acquired cases since May 1 and is aware of over 5,100 additional cases still awaiting the work needed to confirm them, against 249 cases reported nationally by this same point last year. Cyclosporiasis has now been reported in 34 states.
The state figures are far larger. Michigan reported 6,571 cases on July 21, an increase of 423 in a single day, with 102 hospitalizations. Northwest Ohio has reported more than 2,000. Michigan officials have said this is likely the largest outbreak of its kind in United States history, and a state that normally sees roughly fifty cases a year is the epicenter. Michigan also did the exposure work that broke the case: investigators analyzed food histories from 190 of the CDC-confirmed patients and found that 90 percent reported eating iceberg lettuce at Taco Bell.
On July 17, FDA’s traceback investigation identified Taylor Farms de Mexico as the single supplier of the shredded iceberg lettuce served at the implicated restaurants, and the company recalled all iceberg lettuce sourced from central Mexico, product that had gone to 27 states, including Marketside-brand bags sold at Walmart. Then came the detour. On July 18, FDA reported that a sample of Taylor Farms de Mexico shredded iceberg, collected during targeted import surveillance, had tested positive for Cyclospora. On July 19, FDA retracted it, saying that because of the complexity of detecting this parasite, its laboratory experts had re-reviewed the result, concluded the finding did not represent true amplification, and should be treated as a false positive. There are no confirmed positive product samples in this outbreak.
The company moved immediately. Its July 19 statement opens with the line “Today, FDA apologized to us,” and states that FDA has not identified a single positive product test result for Cyclospora. FDA answered publicly the next day: the false-positive sample does not change the basis for the investigation or the overwhelming epidemiological data supporting the recall, and traceback and outbreak data continue to converge on shredded iceberg lettuce from Taylor Farms locations in central Mexico. Health officials said they had offered no apology. FDA’s guidance is unchanged: do not eat the recalled lettuce.
The determination stands, and it stands on what these cases almost always stand on — epidemiology and traceback, not a laboratory hit on the food. That is not a weakness in the case. It is the ordinary condition of a Cyclospora investigation. The incubation period runs a week or more, diagnosis takes weeks longer, and lettuce is perishable. By the time anyone knows to look for the product, the product is gone. The victims ate the evidence in May.
Which is exactly why the two federal documents below are worth reading in full. Both were written after outbreaks in which the pathogen was never recovered from the food, and both show what investigators found when they finally walked into a Taylor Farms facility.
Document One: FDA Environmental Assessment, November 2013
The Cyclospora Outbreak – 631 Sick
In the summer of 2013, cyclosporiasis sickened 631 people across 25 states. Iowa and Nebraska accounted for 239 of them. Epidemiologic and traceback work by the two states, CDC and FDA linked bagged salad mix supplied by Taylor Farms de Mexico, in Doctor Mora, Guanajuato, to Olive Garden and Red Lobster restaurants owned by Darden. A separate 278 cases in Texas were traced to cilantro from Mexico. The cause of 126 illnesses across 22 states was never determined. The implicated salad mix was iceberg lettuce, romaine, green leaf, red cabbage and carrots.
What FDA did
Taylor Farms de Mexico suspended production and shipment to the United States on August 9 and notified FDA on August 12. FDA began work that same day and stayed through August 19, alongside CDC and five Mexican agencies — COFEPRIS, SENASICA, SAGARPA, CESAVEG and the Guanajuato state health authority. The team assessed the processing plant and five ranches: three identified by traceback, plus two more added during the assessment. The plant was not running while they were there.
The timing problem, stated by FDA itself
The last Iowa or Nebraska illness onset was July 2 — more than five weeks before the assessment began. FDA put that sentence in the introduction of its own report.
The findings
No significant source. No significant route of contamination. Employees, restrooms, plumbing, the wastewater treatment plant, harvest crews, field sanitary facilities, irrigation systems and growing fields were each judged “not likely” to be sources — and every one of those conclusions carries the qualifier at the time of the assessment. FDA’s bottom line was that it could not determine how, or at what point in the supply chain, the salad mix became contaminated.
The wash step
The one process point FDA flagged was the washing of the combined salad components in recycled wash water. The firm held free chlorine at a 10 ppm target and used phosphoric acid to keep the pH at or below 7.0. In the same passage, FDA noted that Cyclospora resists halogen-based sanitizers including chlorine, that the infectious dose is estimated at 10 to 100 sporulated oocysts, and that micro-filtration — which can eliminate the parasite from water — is defeated by the turbidity of recycled salad wash water and is therefore not used anywhere in the cut salad industry.
The testing
Roughly 835 product, water and environmental samples were analyzed across FDA, CDC and the company, including more than 269 human fecal samples collected from on-ranch sanitary facilities. All were negative except one: FDA sample 744382, the recycled wash water, came back inconclusive because turbidity inhibited the BAM method. Taylor Farms de Mexico also ran its own parallel assessment covering the processing facility and 30 ranches — 793 samples, all negative.
That distinction matters. “Everything tested negative” is not what the record says. The one sample taken at the exact point FDA identified as the plausible amplification step could not be read at all.
Three details that never made the press
Ranch E. FDA added this ranch to the assessment after reviewing the company’s own irrigation water testing records and finding a fecal coliform positive from drip irrigation tape. The result was 200 MPN/100 mL. It triggered nothing. The firm’s rolling mean-average system did not fire because follow-up samples were negative, and 200 sits just under the company’s own single-sample action level of 235. The crop remained harvestable for salad production.
The prescriptions. FDA tabulated fourteen harvest-crew medical visits inside the outbreak window. Loperamide, an antidiarrheal, was dispensed on June 26, July 3, July 10 and July 24. Trimethoprim-sulfamethoxazole — which FDA notes in the report is effective against Cyclospora — was given on March 6 and twice on August 7, both August doses flagged in the table as near expiration and “chosen for economic reasons.” The company doctor said he had not suspected a parasitic infection. No diagnosis was recorded by the firm for any employee illness.
The carrot ranch. A single ranch in Sonora supplied 100 percent of the carrots in the implicated salad mix. It was never assessed. The harvest season had ended.
The four recommendations
1. Assess the Sonora carrot growing environment before the next carrot harvest.
2. Research whether Cyclospora cayetanensis is a reasonably likely food safety hazard associated with the Guanajuato leafy green growing region — and if it is, have Taylor Farms de Mexico re-evaluate its salad mix processing procedures to control cross-contamination at the washing step.
3. Use the company nurse’s home visits to encourage symptomatic workers to submit stool samples for Cyclosporaanalysis.
4. Fix the ranch handwashing faucets, none of which were hands-free.
Taylor Farms de Mexico resumed production and shipment to the United States on August 25, 2013, after FDA reviewed a sampling plan.
Document Two: FDA Form 483, Taylor Farms Colorado, Inc., issued December 18, 2024
The E. coli Outbreak – 104 Sick
E. coli O157:H7 in slivered onions served on McDonald’s Quarter Pounders. One hundred four confirmed cases across 14 states, 34 hospitalizations, four cases of hemolytic uremic syndrome, and one death in Mesa County, Colorado. Onsets ran from mid-September through October 21, 2024. Taylor Farms recalled the yellow onions on October 22. FDA later stated that the outbreak strain was never found in onion or environmental samples, and that epidemiologic and traceback evidence showed the recalled onions were the likely source. McDonald’s stopped sourcing from the Colorado Springs plant indefinitely and re-supplied roughly 900 restaurants.
The scale of the inspection
Eight days on site: October 28, 29, 30 and 31, and November 4, 5, 7 and 12. Nine FDA investigators signed the form, including a designated National Expert and an Acting Branch Chief. The facility is Taylor Farms Colorado, Inc., 890 N Newport Road, FEI 1783599, and the form was issued to President Christopher M. Hallee. This was not a routine visit.
Observation 1 — the sanitation preventive control was never implemented
The hazard analysis governing slivered onions, dated October 23, 2024, did not identify a preventive control for recontamination by environmental pathogens. The food safety plan that management said covers all processed vegetables at that location — including the onions — carries the product description “Diced Tomatoe.” Slivered onions are exposed to the environment from the peel step through finished packaging, with no lethal treatment anywhere in between.
Production rooms were held at temperatures that would support the growth of Listeria. FDA listed twelve separate non-food-contact locations that tested positive for Listeria spp. between March 20, 2023 and June 14, 2024 — drains, floors, production lines, employee high-traffic areas — plus pooling water near a cutting area that was positive on January 26 and January 30, 2024. The corrective action procedure did not address how recurrence would be prevented. Follow-up testing was not consistently completed; a January 28, 2024 positive produced a follow-up test that was also positive.
And the verification method the plant used to check its own hazard measures the amount of soil on a surface. As FDA put it, that method “is not designed to identify Salmonella and Listeria” — the two organisms the plant’s own food safety plan names as the hazards requiring a sanitation control.
Investigators found biofilm and large amounts of food debris present after the firm’s completed post-operation cleans. On November 5 the firm entered “Pass” at 5:51 a.m. and again at 5:54 a.m. on food contact surfaces that FDA concluded should have been marked “Fail.” A customer complaint from September 23, 2024 documented the consequence: a green-pepper-only finished product arrived containing onions.
Also observed: apron parts inside the handwashing sink, with ties touching the basin and then the hands of employees handling food; employees sanitizing gloved hands after touching insanitary surfaces and returning to ready-to-eat food without ever removing gloves and washing, which management described as common practice; a hose spraying near exposed spin baskets; food particles in the crevices of equipment welds; trapped water and food debris in pitted food contact surfaces; white cutting boards used for ready-to-eat lettuce and celery, severely discolored with deep cracks and crevices; and a torn carrot-line belt with a piece that would come into direct contact with ready-to-eat food before packaging.
The line that ran in every news account: production employees handling ready-to-eat produce and food contact surfaces were not observed using any of the handwashing sinks in the facility.
Observation 2 — the records did not exist
The pre-operation cleanliness check documents for September 24 and October 5, 2024 were missing. The PCQI-trained individual signed off on all preventive control monitoring records in the production packets for both days and did not note the missing records. No corrective action report was generated.
Observation 3 — operations not run under conditions minimizing contamination
Knives stored fully submerged in sanitizing solution were removed and used immediately to chop ready-to-eat produce, without the air-drying step the chemical manufacturer’s own label requires. During production on November 4, an item removed from sanitizing solution measured over 200 ppm against a 50 to 200 ppm range for food contact surfaces, was not air dried, and could transfer that solution to ready-to-eat lettuce. Slicers were sprayed and put back in use within minutes.
And production employees mix their own chemical recipe. It is used as a “reclean step,” scrubbed onto food contact surfaces when checks fail, before the surface is retested. The employee performing the reclean on November 5 said he mixes it routinely. The firm could not locate the source of the recipe, and management could not provide a manufacturer label or any chemical representative’s statement that the mixture was designed for that use.
The classification
FDA classified the inspection VAI — voluntary action indicated — meaning no administrative or regulatory action followed. Taylor Farms has cited that classification publicly as vindication, alongside the observation that no illnesses were linked to the specific conditions FDA wrote down. Both statements are accurate. Neither addresses what the nine investigators saw.
What the Two Documents Show Together
Both were opened because people were already sick. Neither is a routine inspection. In 2013, FDA arrived five weeks after the last illness onset, at a plant that had voluntarily stopped running. In 2024, FDA arrived a week after the recall. Nine investigators and eight days in Colorado Springs; a five-agency binational team and eight days in Guanajuato. That level of federal attention is what a body count buys.
Neither found the pathogen. No Cyclospora in 2013. No outbreak-strain E. coli in 2024. Both times the case rested on epidemiology and traceback. Both times the company said so publicly. That is the identical posture Taylor Farms holds today. It was not an exoneration in 2013, it was not one in 2024, and it is not one now.
What each document did find was a control the company had written down and had not operated. In 2013: a wash step FDA identified as the plausible amplification point, running chlorine against an organism FDA said in the same paragraph resists chlorine, and an irrigation trigger set at 235 that a 200 sailed straight under. In 2024: a hazard analysis that omitted the hazard, a food safety plan labeled for the wrong vegetable, twenty months of Listeria positives with corrective actions that did not correct, a verification method incapable of detecting the organisms it was meant to verify, and pre-operation checks signed “Pass” on visibly dirty equipment. Different pathogen, different hemisphere, different regulation — the same failure mode. The paper system exists. The physical system does not follow it.
And the 2013 recommendation is the one that matters this month
FDA told Taylor Farms de Mexico to determine whether Cyclospora is a reasonably likely food safety hazard associated with the Guanajuato leafy green growing region, and if it is, to re-evaluate the wash step. That was November 2013.
It is now July 2026. Same subsidiary. Same Guanajuato region. Same parasite. Iceberg lettuce. And by the state health departments’ own counts, more than ten thousand people.
There is one question someone should answer. Was that research ever done, and what did it conclude? If it was done and it found that Cyclospora is a reasonably likely hazard in central Mexico, then the wash step should have been redesigned in 2014. If it was never done, the company left FDA’s central recommendation sitting on the table for thirteen years, and this outbreak is the second invoice for it.