How the CDC Stopped Tracking Raw Milk's Germs as Sales Soared
As raw milk demand climbs and states loosen the rules, the CDC made six of its eight tracked food-borne germs optional to report, including those raw milk carries.
Distrust, wellness, and the post-2020 raw-milk boom.
The 2026 raw-milk outbreaks are already here.
How pasteurization won over the objections of raw-milk advocates.
Safety claims rest on shaky numbers.
CDC’s FoodNet went from eight germs to two.
Who can actually fix this.
Distrust, wellness, and the post-2020 raw-milk boom
Erling Haaland drinks raw, unpasteurized milk and calls it a superfood. One of the biggest stars at this summer’s World Cup credits it for how he feels and trains. Millions of people online make the same case for raw milk, and more of them can buy it every month. Demand has climbed since 2020, driven by distrust of public health after COVID and a wellness movement that frames raw milk as “natural,” a “cure-all,” and a way to reject expert advice. States keep loosening the rules. Oklahoma just raised its legal sales cap fifteen-fold, in the middle of an outbreak.
The 2026 raw-milk outbreaks are already here.
This spring, nearly 60 people in Idaho got sick from raw milk, most of them from Campylobacter. There have been more in Louisiana and California. A New Mexico advisory followed the death of a newborn whose mother drank raw milk contaminated with Listeria in pregnancy. A California producer recalled raw cheddar after an E. coli O157:H7 outbreak that sent a child into kidney failure. Ounce for ounce, raw dairy causes about 840 times more illness than pasteurized.
How pasteurization won over the objections of raw-milk advocates.
In 1891, about a quarter of the babies born in New York City died before their first birthday, and dirty raw milk was a leading cause: it carried tuberculosis, typhoid, scarlet fever, and diphtheria. Louis Pasteur had shown in the 1860s that gentle heat killed the microbes that spoiled wine and beer, and reformers spent decades pressing to apply the idea to milk. One of them, the Macy’s co-owner Nathan Straus, had lost children to milk-borne illness. He opened free pasteurized-milk stations in poor neighborhoods, and of about 20,000 children he fed over four years, only six died (a death rate of ~0.03% versus 24% city-wide infant mortality). Cities mandated pasteurization only after a long fight, and the people fighting it were “certified raw milk” advocates who called heating milk unnecessary. It is the same argument, almost word for word, being made today.
Safety claims rest on shaky numbers.
The industry says it’s getting safer. The Raw Milk Institute, which lobbies for the product, cites a study showing that raw milk outbreak rates are falling even as consumption rises. The Institute’s founder also owns Raw Farm, the producer of the cheddar implicated in the E. coli O157:H7 outbreak. Regulators have linked the company to 233 illness cases across two decades.
CDC’s FoodNet went from eight germs to two.
But a falling rate depends on someone doing the counting, and on July 1, 2025, the federal government stopped requiring most of it. The Centers for Disease Control and Prevention cut its flagship food safety network, FoodNet, from eight tracked germs to two. Salmonella and Shiga toxin-producing E. coli stayed mandatory, while Campylobacter, Cyclospora, Listeria, Shigella, Vibrio, and Yersinia became optional.
Two of the germs now optional are the ones raw milk carries. The Idaho outbreak was caused by Campylobacter, and testing is now optional. Testing for Listeria, which killed the newborn and is dangerous in pregnancy, is now optional. The one raw milk pathogen still mandatory is STEC, the toxin behind the cheddar recall. The government made it harder to monitor germs in raw milk at the same time that states made raw milk easier to sell.
It helps to know what FoodNet does. Since 1995, when it was established after the Jack-in-the-Box E. coli deaths, it has run active surveillance across about 15% of the country through 10 state sites, auditing labs to ensure every diagnosed case is reported. That is the source of these stats: 48 million Americans are sickened by food each year, 128,000 are hospitalized, and 3,000 die.
2 of the 6 are among the deadliest germs we track. Listeria is the third-leading cause of death from food-borne illness, and among people sick enough to be diagnosed, roughly 15% to 20% die. Campylobacter sickens about 1.5 million Americans a year, and that’s an underestimate. For some germs, the gap is huge. The CDC estimates that there are about 130 undiagnosed Vibrio infections for every one it records. FoodNet surveillance was how the CDC corrected for that under-counting.
“States can continue to monitor” means the floor is now a patchwork: Minnesota and Maryland say they will keep all eight; Colorado says it may cut back. That is the same logic as the raw milk law itself: some states are careful, some states are blank, your protection is decided by where you live.
Cutting surveillance makes harm invisible without reducing it. The number the lobby uses to argue that raw milk is getting safer is produced by the exact system now being defunded.
Who can actually fix this.
None of this can be fixed in the dairy aisle. A shopper can’t weigh a risk that has been removed from the data. The levers sit with the people who pulled them: HHS leadership, which made the call and cited funding; congressional appropriators, who decide whether FoodNet is worth its line item; and the ten state health departments now choosing, one by one, whether to backstop a federal retreat.
A raw milk outbreak in a state that has stopped counting is a never event: the kind of preventable harm that should never happen.


