Two years after it emerged, ‘cow flu’ is still circulating

After 2 years, the U.S. outbreak of the H5N1 influenza virus in cattle appears to be waning, easing fears that the virus could cause long-lasting damage to the dairy industry or mutate into a form that could cause a human pandemic. The last new detection of an affected herd occurred on 13 December 2025 at a Wisconsin farm, according to the website of the U.S. Department of Agriculture (USDA). Data from a USDA program that tests milk for the virus suggest 16 of the 19 affected states have gotten rid of the virus, which originated in wild birds. But it could still bounce back, and efforts to eliminate it entirely face formidable challenges. How it spreads between cattle remains unclear. And although candidate vaccines look promising in early tests, farmers and the government may be reluctant to embrace them. H5N1 continues to circulate on farms in California and Idaho, those states’ agriculture departments told Science. Texas remains “affected” in USDA’s latest update, even though it has not had a detection since May 2025, because it has not complied with the National Milk Testing Strategy requirement that it sample all silos at milk-processing plants. In other states the virus may simply have escaped detection. The first cases of H5N1 in dairy cattle were discovered in Texas in March 2024. The virus—a variant of H5N1 known as 2.3.4.4b—has been killing poultry and wild birds worldwide since 2020. It rarely kills cows but it thickens their milk, turns it yellow, and leads to steep drops in production. Early in the outbreak, milking machines—which can be used on hundreds of cows every day—were fingered as an important conduit for the virus. The fact that infected cows had inflamed udders and milk from affected farms had high levels of the virus seemed to support the suspicion. But research has revealed a far more complicated picture. “It seems like there are a variety of different ways that the virus is transmitted,” says epidemiologist Jason Lombard, a veterinarian who worked at USDA for 20 years and now is at Colorado State University. A growing number of scientists think the virus readily drifts on the wind from farm to farm and cow to cow. Contaminated waste milk fed to calves could infect them. One study found the virus in semen from a bull, which could lead to infections in cows, too. Even flies may transmit H5N1. “I don’t think we’re 100% sure about anything when it comes to figuring out how this virus is getting around,” says virologist Richard Webby, an avian influenza specialist at St. Jude Children’s Research Hospital. As the outbreak dwindles, scientists have fewer opportunities for field studies that could provide answers. And some complain that funding has become scarce and USDA has been slow to share its data. “I’m frustrated because I feel like we should know more,” says veterinarian Andrew Bowman of Ohio State University. Still, control measures have made major inroads. After recognizing that cow shipments spread the virus, USDA required that lactating dairy cattle test negative before they could be moved between states. To pinpoint infected farms, USDA also launched the milk-testing program. It required labs and vets to report positive tests and farmers to provide information about the movement of animals in and out of affected herds. State-mandated quarantining of infected herds also helped, as did natural immunity in herds that saw an outbreak. “Most herds and affected states have been able to eliminate the virus,” USDA told Science in a statement. (The agency denied requests to interview its scientists.) But the virus remains ensconced in some farms. Veterinarian Edith Marshall, an epidemiologist with the California Department of Food and Agriculture, says a few have had cows that remain infected for more than 1 year. These cows could act as “superspreaders” of the virus to the new calves and heifers, which typically replace one in every three adult cows in a dairy herd every year. Infections could also result from infected cows that show no signs of disease and never are tested, or if tests produce false negatives. Marshall, Lombard, and Webby teamed up with a California veterinarian, Blaine Melody of the Lander Veterinary Clinic, to study in detail how the virus is transmitted. In a preprint posted last year, they looked closely at the role of milking machines at 14 California dairy farms. The milking equipment harbored the virus, but lab studies by Bowman and USDA that attempted to transmit it to cows through contaminated machinery have failed. What’s more, a cow udder has four teats, and many only have the virus in one “quarter.” Milking machines have four “teat cups” that milk the same quarter on each cow, but the study of the California dairies found no pattern between the infected quarters in a given herd. “If it was the milking equipment, I would think that we’d see the same quarters infected,” says co-author Seema Lakdawala, a virologist at Emory University. Data from several studies hint at other transmission routes. Air samplers at the California farms readily detected the virus, supporting the idea that the virus transmits through the air. And in a separate long-term study at 18 farms in California’s Central Valley, Melody found provocative, as-yet-unpublished evidence suggesting the wind can spread the viruses long distances. Dairy farms in the southern part of the Central Valley had an explosion of the virus in the early fall of 2024, whereas those in the north were spared. Then in late November, a bomb cyclone walloped the state, sending strong southern winds across the northern Central Valley. Within 4 days, all 18 farms had an outbreak. Webby is now sequencing virus samples from those farms, which might confirm that their infections had a common origin. Previous research has suggested the wind can give flu viruses wings. A study from the Netherlands showed it played an important role in the spread of a deadly subtype named H7N7 between poultry farms in 2003, and a 2025 report from the Czech Republic contends the wind moved H5N1 between poultry farms separated by 8 kilometers. “The head scratcher we all still have is how does it get from a particle blown in the wind into the udder of a cow?” Webby asks. But Melody is convinced cows’ noses are a main route for infection, even though nasal swabs often test negative. Flies may also help transmit the virus. Biting flies are known to transmit bacteria that cause mastitis between cows, and last year, Lombard and Melody reported that cows had H5N1 in their bloodstreams. “Could flies be biting the cows and injecting virus?” Lombard asks. Marshall and colleagues have found the virus on house flies and blow flies. Those species don’t bite, but they regurgitate what’s in their guts, another potential route of transmission, the researchers suggested in a November 2025 preprint. “If a fly is vomiting on the teat end of a cow, is there enough viral particle in there?” Marshall asks. “Waste milk” from infected cows could also be a route of spread, says veterinarian Richard Pereira, an epidemiologist at the University of California, Davis. Only about half of large farms and fewer than 1% of medium and small ones pasteurize waste milk, which is often fed to calves or dumped, a potential threat to humans and pets. Pereira has developed a method to inactivate the virus in waste milk by adding citric acid to it, which is far cheaper than pasteurization. Even if H5N1 were eliminated from U.S. dairy herds, it could jump back in, scientists warn. The virus now circulating in cows, a genotype called B3.13, stems from a single introduction into cattle that likely occurred in the fall of 2023. But results from milk testing showed that another genotype, D1.1, jumped into cattle from birds on three occasions, quickly dying out each time. An unknown variant seems to have made the jump in the Netherlands, where researchers recently found antibodies against H5N1 in cows, indicating they were infected at some point. Some researchers say vaccination is the best way to bring the current outbreak to an end and prevent future ones. “The goal has to be to use a vaccine in conjunction with testing and all the other things in place right now to really try to stamp this thing out,” says viral immunologist Scott Hensley of the University of Pennsylvania. Hensley’s team has one of several promising vaccine candidates under development. Earlier this month, the group posted results from a trial of a messenger RNA (mRNA) vaccine that triggers production of B3.13’s hemagglutinin surface protein. The study found animals that were vaccinated and then exposed to the virus had reduced disease and produced just as much milk, with 1000-fold lower viral levels in it. Whether USDA will allow use of H5N1 vaccines in dairy cattle—and whether the price will be attractive to farmers—is a huge question. H5N1 vaccines exist for poultry, but USDA has never given them a green light, mainly because of concerns that vaccination could harm the export market. And because of widespread suspicion of mRNA vaccines in the wake of the COVID-19 pandemic, many states are considering banning their use for livestock as well. “It’s amazing to me,” Hensley says. Pereira sees the limited spread of the virus right now as a lucky grace period that should be used to speed vaccine development and use every existing measure to prevent a resurgence. “Instead of, ‘Oh, no, it’s OK now, we can ignore it,’ we should realize we’ve been given time,” he says. “Let’s use it wisely.”