A raw milk dairy that Washington State health investigators identified as the most likely source of four E. coli infections earlier this year is still selling its milk because a follow-up sample collected weeks after the illnesses tested negative and no recall was recommended.
The situation, documented on the state's outbreak investigation page, is settled rather than new. The illnesses began in late spring, the department published its findings in early summer, and no product action followed. For anyone in the Pacific Northwest who buys raw milk, the practical question is a live one today: the state named a likely source and left the product on the shelf.
That gap between an epidemiologic finding and a regulatory action is the story here, and it is a recurring feature of raw milk investigations rather than a one-off.
A Likely Source Without a Confirming Sample
The department's conclusion was carefully worded. Based on the available epidemiologic information, it said, Dungeness Valley Creamery raw milk is considered the most likely source of exposure among the cases.
What it did not have was a positive product sample. A routine milk sample collected by the Washington State Department of Agriculture on June 16 tested negative for Shiga toxin-producing E. coli. Because of that negative result, the department said no recall was recommended. The agriculture department tests milk from all retail raw milk operations monthly.
Both facts can be true at once, and understanding why matters for how readers should weigh this. Contamination in raw milk is intermittent. It depends on what is shed by which animal on which day, and a sample drawn weeks after the illnesses occurred tests the milk in the bottle that day, not the milk people drank in late May. A negative result does not retroactively clear an earlier batch. It also does not prove the earlier batch was contaminated.
This should be read as an epidemiologic association rather than laboratory-confirmed contamination of a specific product.
How the Four Cases Were Linked
One detail strengthens the investigation. All four people were infected with the same highly related strain of Shiga toxin-producing E. coli, with illness onsets between May 28 and June 4. That genetic match is what makes these four infections an outbreak rather than a coincidence.
The exposure histories are less complete. Two people reported drinking Dungeness Valley Creamery raw milk shortly before becoming ill: an Oregon resident who purchased the milk on May 25 and a Snohomish County, Washington resident who purchased it on May 28.
The third was a King County resident who reported consuming raw milk before becoming ill but declined to identify the brand. The fourth was a Texas resident who tested positive while visiting Washington State and could not be interviewed.
So two of four cases name the dairy directly, one names raw milk without a brand, and one could not be reached. The department described sharing the information as part of its routine outbreak response, framed around consumer awareness. None of the four was hospitalized, and none died. The state lists the outbreak as over.
The dairy, based in Sequim, has appeared in the state's food safety record before. An outbreak in 2009 was linked to raw whole milk from the creamery under previous ownership and involved three cases, and in 2013 state agriculture officials warned consumers away from several of its raw products after routine sampling found toxin-producing E. coli in a cream sample. Trade coverage has also documented the dairy's earlier contamination record, including later recalls. That history is relevant context, not evidence about the 2026 cases.
The Gap Between Epidemiology and Recall Authority
The pattern visible here is not unique to one dairy or one state, and it has become a live national question.
Raw milk investigations frequently end without a recall for the same structural reason. Illnesses surface weeks after exposure, by which time the implicated batch has been consumed and the next batch tests clean. Regulators generally need a positive product sample or a firm's cooperation to move a product off the market, and voluntary recall remains the default mechanism. The contrast is visible in the same state: when Washington investigated an outbreak tied to aged raw milk cheese, product samples matched the strains found in patients, and a recall followed.
That default has drawn congressional attention to voluntary recall procedures after a larger raw dairy outbreak in which the producer initially declined to recall implicated cheese before doing so voluntarily.
Washington permits licensed raw milk sales, subject to inspection and testing requirements, and the state notes plainly that licensing does not entirely remove the risk of contamination. Nothing about this investigation changed that status, and consumers can still buy the product.
Who Should Avoid Raw Milk Entirely
The medical consensus here is not contested, and this is one of the cases where balance would be false.
Pasteurization reliably kills Campylobacter, E. coli, Listeria, and Salmonella without meaningfully changing milk's nutritional composition. The American Academy of Pediatrics, the CDC, the FDA, and the American Medical Association all recommend against consuming raw milk and specifically against giving it to children.
Risk is concentrated in children under 5, pregnant people, older adults, and anyone with a weakened immune system, and state guidance on the risks of drinking raw milk says those groups should not drink it at all. In children, Shiga toxin-producing E. coli infection can progress to hemolytic uremic syndrome, a condition that destroys red blood cells and can cause acute kidney failure.
Symptoms of STEC infection typically begin three to four days after exposure but can take up to nine days, and include severe stomach cramps, diarrhea that is often bloody, and vomiting. Some infected people have no symptoms at all. Anyone with bloody diarrhea, diarrhea lasting three days or more, or signs of dehydration should call a clinician right away. Reduced urination, extreme fatigue, or paleness in a child after a diarrheal illness needs urgent evaluation. Antibiotics are generally avoided in suspected STEC infection because they may increase the risk of kidney complications, which is a decision for a clinician.
Households that buy raw milk should know that no producer, however careful, can test its way to safety on every batch, and that state licensing does not eliminate the underlying risk. Prior MedicalDaily reporting laid out what pasteurization actually prevents across recent outbreaks. What remains unknown is whether further sampling has occurred since June and whether federal recall authority will change. MedicalDaily will follow state sampling records and any legislative movement.
Key Questions Answered
What did Washington State find? Health investigators concluded that raw milk from Dungeness Valley Creamery was the most likely source of exposure among four Shiga toxin-producing E. coli cases.
Why was there no recall? A routine agriculture department sample collected in mid-June tested negative for STEC, and the state said no recall was recommended because of that result.
Does a negative sample mean the milk was safe? Not necessarily. Contamination in raw milk is intermittent, and a sample taken weeks later tests that day's milk rather than the batch people drank.
Were all four cases linked to this dairy? All four shared the same closely related strain, but only two named the dairy. One reported raw milk without a brand, and one could not be interviewed.
Is the product still for sale? Yes. Washington permits licensed raw milk sales, and no recall was issued.
Who is most at risk from raw milk? Children under 5, pregnant people, older adults and people with weakened immune systems. Children can develop hemolytic uremic syndrome, a form of kidney failure.
What symptoms should someone watch for? Severe cramps, diarrhea that is often bloody, and vomiting, typically three to four days after exposure. Bloody diarrhea or diarrhea lasting three days or more warrants calling a clinician.
