All summer I kept a running count of the 2026 Cyclospora season from CDC, FDA and the fifty state health departments. A month on from my last straight count, three questions arrive most often: where does it stand, which illnesses are tied to Taylor Farms, and are people still getting sick. Here are the answers as of Thursday, October 15.
The short answers.
The outbreak CDC and FDA tied to shredded iceberg lettuce from Taylor Farms de Mexico closed at 12,883 illnesses, 570 hospitalizations and two deaths in 21 states.
CDC’s season’s count of laboratory-confirmed domestic cases stopped on September 15 at 19,883, with 1,064 hospitalizations, and the agency said it knew of 36,242 people in all. CDC’s weekly notifiable-disease tables, which never stopped, carried 32,286 cyclosporiasis cases for 2026 through October 3, against 4,042 at the same point in 2025. New illnesses ended in August.
The outbreak number: closed at 12,883
CDC declared the outbreak over on September 11. FDA closed its investigation on October 2 after finding the parasite in a tank of outgoing wastewater at the Taylor Farms de Mexico processing plant and in a drainage ditch at one of the iceberg growers in its traceback. FDA says the two environmental positives add to strong epidemiological data and traceback convergence, that genotyping of both samples is complete, and that it is still analyzing how they relate to the genotyping of patient specimens. Thirteen days on, that comparison has not been published. The last illness onset FDA lists is August 17. FDA also promised an Outbreak Investigation Report and ten activities before the 2027 growing season; I went through the ten on October 7.
The season number: 19,883, on a page that stopped on September 15
CDC’s Cyclospora Case Data page counts every laboratory-confirmed case acquired in the United States with symptoms starting during the May 1 to August 31 season, tied to the lettuce or not. Its last update, September 15: 19,883 confirmed cases, 1,064 hospitalizations, two deaths, 49 states and the District of Columbia. Another 9,765 confirmed cases were still being sorted into domestic or travel-related, and CDC said it was aware of at least 6,594 more that were not yet laboratory-confirmed or were waiting on an interview. That is 36,242 people. The same update said CDC would stop weekly updates because recent infections had significantly declined and would post periodically. Thirty days later the page still says September 15.
The illnesses linked to Taylor Farms, and the ones that are not
Linked means something specific. CDC’s definition, carried on every update from August 5 on, is people who reported exposure to processed iceberg lettuce from Taylor Farms de Mexico served or purchased at various locations, including people who previously reported eating at Taco Bell. Attribution is reported exposure in an interview. The count began on July 16 as 1,644 Taco Bell patrons in five states, ninety percent of whom, in Michigan’s ingredient analysis, had eaten iceberg lettuce; it widened on August 5 to anyone who reported the recalled lettuce, and it closed at 12,883. CDC’s map assigns 5,762 of them to Michigan, 2,871 to Ohio, 1,071 to Missouri and 870 to Indiana. CDC never published how many ate at Taco Bell and how many ate the lettuce somewhere else; the one place outside Taco Bell where a state tied cases to a named Taylor Farms product was the cafeteria at Cheshire Medical Center in Keene, New Hampshire.
Everything else is not linked, and it comes in layers. Of the 19,883 confirmed domestic season cases, 7,000 are outside the outbreak. FDA’s outbreak table accounts for 313 of them, in four Cyclospora files that ended with no food identified. The rest were never in any public investigation, and they sit where the lettuce did not go. New York has 1,121 cases in the federal table and none on the lettuce map, Florida 543 and none, California 428 and none, North Carolina 930 and three. New York traced an April cluster to cilantro at a Brooklyn restaurant; North Carolina pointed to cilantro and parsley. Beyond those sit 9,765 confirmed cases not yet sorted into domestic or travel, 2,078 confirmed travel cases, and 6,594 reports not yet laboratory confirmed. The largest unlinked block is Michigan, which counts every confirmed and probable case reported during the outbreak period, lettuce or not: 14,718, of which the lettuce map takes 5,762. Michigan interviewed more than 2,900 patients, found many who never ate at Taco Bell or any restaurant, and still saw lettuce turn up often, which the state reads as contamination upstream of any one restaurant.
The count that never stopped: 32,286 through October 3
There is a third federal number, and it is the one almost nobody reads. Every week CDC publishes the National Notifiable Diseases Surveillance System tables, the provisional case notifications state health departments send to Atlanta for every reportable disease, cyclosporiasis included. They run from the first week of January, they include travel cases, and the case page warns that the two counts may not match. They are also the only federal cyclosporiasis count still published on a schedule. The table for the week ending October 3 shows 32,286 cases reported to CDC in 2026 against 4,042 at the same week in 2025, eight times last year; the full year 2025ended at 3,952 and 2024 at 4,210. In the week ending October 3 alone, states reported 289 cases, 244 of them from Michigan. The October 10 table had not reached CDC Stacks when I checked; an independent tracker that pulls the same dataset reads it at 33,281, with Michigan at 15,575.
Here are all fifty states, from two CDC sources: the number of cases each state reported to CDC in 2026 through October 3, the same week of 2025, this year’s count as a multiple of last year’s, and the residents of each state the outbreak map assigns to the lettuce. New York includes New York City, which CDC reports separately. The seven jurisdictions that have reported no 2026 cases and the three where cyclosporiasis is not reportable are listed last; Delaware’s 2025 figure was unavailable to CDC. All figures are provisional.
| State | 2026 through Oct. 3 | Same week, 2025 | Multiple of 2025 | Lettuce outbreak, CDC map |
| Michigan | 14,660 | 55 | 267x | 5,762 |
| Ohio | 6,527 | 72 | 91x | 2,871 |
| Illinois | 1,534 | 299 | 5.1x | 292 |
| New York (incl. NYC) | 1,121 | 669 | 1.7x | 0 |
| North Carolina | 930 | 296 | 3.1x | 3 |
| Kansas | 857 | 31 | 28x | 373 |
| Kentucky | 837 | 45 | 19x | 366 |
| Oklahoma | 572 | 41 | 14x | 269 |
| Florida | 543 | 191 | 2.8x | 0 |
| California | 428 | 238 | 1.8x | 0 |
| Iowa | 345 | 71 | 4.9x | 138 |
| Nebraska | 345 | 51 | 6.8x | 207 |
| West Virginia | 317 | 7 | 45x | 170 |
| Wisconsin | 303 | 65 | 4.7x | 0 |
| New Jersey | 301 | 175 | 1.7x | 0 |
| Colorado | 279 | 201 | 1.4x | 0 |
| Massachusetts | 262 | 72 | 3.6x | 18 |
| Virginia | 254 | 107 | 2.4x | 12 |
| Arkansas | 245 | 13 | 19x | 87 |
| Georgia | 238 | 111 | 2.1x | 3 |
| Tennessee | 227 | 34 | 6.7x | 7 |
| Maryland | 206 | 121 | 1.7x | 0 |
| Louisiana | 176 | 97 | 1.8x | 0 |
| Alabama | 132 | 21 | 6.3x | 0 |
| Connecticut | 113 | 43 | 2.6x | 0 |
| Arizona | 109 | 44 | 2.5x | 0 |
| South Carolina | 95 | 44 | 2.2x | 0 |
| Utah | 67 | 38 | 1.8x | 0 |
| Maine | 62 | 4 | 16x | 26 |
| Oregon | 56 | 23 | 2.4x | 0 |
| Indiana | 35 | 27 | 1.3x | 870 |
| Nevada | 20 | 9 | 2.2x | 0 |
| Wyoming | 19 | 1 | 19x | 0 |
| Rhode Island | 17 | 6 | 2.8x | 0 |
| New Mexico | 12 | 9 | 1.3x | 0 |
| Alaska | 11 | 5 | 2.2x | 0 |
| Texas | 10 | 493 | 0.02x | 24 |
| New Hampshire | 5 | 7 | 0.7x | 43 |
| North Dakota | 5 | 4 | 1.3x | 0 |
| Vermont | 5 | 2 | 2.5x | 0 |
| Montana | 2 | 9 | 0.2x | 0 |
| Puerto Rico | 1 | — | — | 0 |
| Delaware | none reported | n/a | — | 0 |
| District of Columbia | none reported | 5 | — | 0 |
| Hawaii | none reported | 1 | — | 0 |
| Minnesota | none reported | 65 | — | 0 |
| Missouri | none reported | 58 | — | 1,071 |
| South Dakota | none reported | 3 | — | 0 |
| Washington | none reported | 59 | — | 0 |
| Idaho | not reportable | — | — | 0 |
| Mississippi | not reportable | — | — | 0 |
| Pennsylvania | not reportable | — | — | 271 |
| United States, total | 32,286 | 4,042 | 8.0x | 12,883 |
The states the federal table cannot see
That table is a floor, and the holes in it are the story. Missouri shows no cyclosporiasis cases reported to CDC in 2026 at all, in a year its health department had counted 1,950 by mid-August and the lettuce map carries 1,071. Indiana shows 35 against 870 on the map and 2,104 on its own page when it closed the count on September 3. Pennsylvania shows N, not reportable, because it reports voluntarily. Texas, whose own dashboard passed 300 in August, shows 10. New Hampshire counted 69 and shows 5. Minnesota and Washington show none. Missouri and Indiana alone counted more than 4,000 cases, and the one federal count still running carries 35 of them. Whatever 32,286 is, it is low.
Michigan: 14,718, last posted September 3
Michigan’s last posted figure is 14,718 cases, 366 hospitalizations and two deaths as of September 3. The page then promised a first Thursday-of-the-month cadence, which made October 1 the first new number since the season closed. I cannot find one. CDC’s weekly table has Michigan at 14,660, and if the October 10 reading holds, the federal count of Michigan residents has passed the state’s own. On September 3 I added up what the fifty states were publishing and got 33,457. The federal table is now within a few hundred of that, from below. The states were not inflating anything. They were early.
Are people still getting sick? The illnesses stopped in August. The paperwork did not.
The counts are still moving. The illnesses are not. Every onset series ends in August: CDC’s season page lists the latest domestic onset as August 30; the outbreak’s last onset was August 17, and CDC said that after more than 1,000 infections a day at the July peak, August averaged fewer than two; Michigan’s own curve shows 21 onsets the week of August 9, none the week of August 16, one the week of August 23. The live signal is the current-week column of the federal table with Michigan and Ohio removed: 289 minus 244 minus 28 leaves 17 cases from the other 48 states, against six in the same week of 2025 and 13 the week after. That is a normal October. FDA has opened no new Cyclospora file since August 5.
The reports keep arriving for four reasons. CDC says it takes about six weeks from symptom onset to a case reaching Atlanta, and its page warns that this season’s volume means earlier cases will keep coming. States transmit in batches; Michigan’s largest single week this year was 4,479. Cases move between buckets as interviews sort them into domestic or travel. And travel cases arrive year-round. The infections stopped because the vehicle did: distribution ran June 29 to July 16, the recall came July 17, shredded lettuce lasts days, and one incubation period past the last shipment is the end of July, which is where the curve collapses. Still sick is not newly infected. Untreated cyclosporiasis can run a month or longer and relapses are common, so July patients were back in doctors’ offices in September. Same cases counted once.
How many people were actually sick
Every number above counts people who went to a doctor, had a stool test that looked for this parasite, and had the result reported. Most sick people do none of those things; CDC’s own case page says the true number is likely higher than the number reported, because some people recover without medical care and are never tested. The federal estimate of how much is missed comes from the 2011 CDC burden study by Scallan and colleagues, still the reference: about 83 illnesses for every laboratory-confirmed Cyclospora case. Applied as CDC built it, to laboratory-confirmed domestic cases, the arithmetic is 12,883 times 83.1, about 1.07 million illnesses in the lettuce outbreak alone, and 19,883 times 83.1, about 1.65 million for the season, before the 9,765 confirmed cases CDC had not finished sorting. For perspective, the same study estimated 11,407 domestically acquired foodborne Cyclospora illnesses in a normal year for the whole country, counting everyone who never saw a doctor. This summer’s confirmed cases alone ran to 1.7 times that.
Two caveats cut the other way. The multiplier was built on surveillance from the 2000s, when a laboratory looked for Cyclospora only if a doctor asked; the multiplex stool panels now common in hospitals find it without being asked, and this summer’s publicity sent people for testing who would otherwise have stayed home. Both push the true multiplier below 83. The study’s own credible interval ran from no undercount at all to more than 270 times. The honest statement is a range: at least the 36,242 people CDC says it knows about, and, by the only federal method on the books, something above a million. No agency has published an estimate for this season. CDC has the method and the data. It should publish the number.
What to watch
Nothing I listed on October 2 has moved: no genotype comparison, no Outbreak Investigation Report, no hearing, no answer to the fifteen letters. Watch the NNDSS table each week, FDA’s report, Michigan’s page on the first Thursday of November, and the periodic CDC update promised on September 15. When Michigan posts, I will publish the fifty-state closing table, each state’s last posted count beside CDC’s weekly figure.
The season is over. The outbreak is closed. The investigation is closed. The counting is not. Thirty-two thousand two hundred eighty-six people have a laboratory report with their name on it. Two are dead. Twelve thousand eight hundred eighty-three have a food named. Seven thousand do not. I will keep adding it up.
Recent Comments