Raw Fish and Dogs: Parasites, Thiamine Deficiency, and What Vets Actually Say

TL;DR

• Salmon poisoning disease is the serious one. Raw salmonids in the Pacific Northwest can carry it. Untreated, up to 90% of affected dogs die within 7 to 10 days. It is very treatable if caught.
• Thiaminase destroys vitamin B1. Raw freshwater fish and raw clams carry it. This is a chronic-diet problem, not a one-meal one.
• Bacteria are real but usually manageable. An FDA study found Salmonella and Listeria in commercial raw pet food.
• Cooking removes all three risks completely. Heat kills the parasite, denatures thiaminase, and kills the bacteria.
• Freezing helps with parasites, not with thiaminase or bacteria. It is a partial fix, not an equivalent one.
• Raw feeding is a genuine disagreement. Regulators and most veterinary bodies advise against it; experienced raw feeders manage the risks deliberately. Both positions are laid out below.

Risk one: salmon poisoning disease

This is the one that justifies the caution. Salmon poisoning disease is caused by Neorickettsia helminthoeca, a bacterium that lives inside a fluke, which is a parasitic flatworm called Nanophyetus salmincola. The fluke lives in salmonid fish. When a dog eats raw or undercooked infected fish, it takes in the fluke, and the fluke delivers the bacterium.

The name is misleading in two ways. It is not poisoning, it is an infection. And it is not limited to salmon: trout and other salmonids carry the fluke too. Steelhead and Pacific giant salamanders have also been implicated.

The Merck Veterinary Manual places its US range in coastal areas of the Pacific Northwest, roughly northern California to Puget Sound, and inland along the rivers those fish migrate up. Cases have also been reported in southern California and Brazil. If you are outside that range and your fish came from a grocery store, this specific risk is low.

What salmon poisoning looks like, and what happens next

Merck describes an abrupt onset roughly 7 days after the meal. Fever appears in more than 70% of cases. So does generalized swelling of the lymph nodes. Vomiting and diarrhea are typical, and the diarrhea may contain blood or turn black and tarry. Dehydration and severe weight loss follow. Thrombocytopenia, a drop in platelets, occurs in up to 90% of cases.

Without treatment, up to 90% of affected dogs die within 7 to 10 days of signs appearing. That mortality figure is the reason this page exists.

With treatment, the outlook changes completely. Merck gives the treatment of choice as doxycycline at 5 mg/kg by mouth every 12 hours for a minimum of 7 days, with praziquantel to clear the fluke itself. Diagnosis uses clinical signs plus history plus a fecal exam or PCR. This is a disease where telling your vet what your dog ate is worth more than any test, because the history points straight at it.

• Fever, in over 70% of cases
• Swollen lymph nodes, in over 70% of cases
• Vomiting and diarrhea, sometimes with blood or black tarry stool
• Marked weight loss and dehydration
• Nasal or eye discharge
• Onset about 7 days after eating raw fish, occasionally up to 33 days

Risk two: thiaminase and vitamin B1 deficiency

Thiaminase is an enzyme that breaks down thiamine, which is vitamin B1. Some raw fish contain it, and if a dog's diet is built on those fish, the enzyme destroys the thiamine in the food faster than the dog can absorb it.

The species distinction matters, and most pages get it wrong. The Merck Veterinary Manual specifies raw freshwater fish and raw clams as the thiaminase sources. It is not a property of all fish, and it is not a property of any cooked fish.

Merck is explicit that thiaminase is destroyed by cooking, and notes that the heat used in canning is enough on its own. That is why canned fish carries no thiaminase risk at all, even though it is not something you cooked yourself.

Thiamine deficiency in dogs shows up as loss of appetite, depression, weakness in the hind limbs, seizures, coma and death. Merck notes it is more common in cats than dogs, and that it does not develop in animals fed properly formulated complete diets. This is a risk of a fish-heavy homemade raw diet, not of an occasional raw fish treat.

Risk three: bacteria

Raw fish carries the same bacterial risks as any raw animal protein. The ASPCA Animal Poison Control Center's guidance on people foods lists raw meat, raw eggs and raw fish together, on the basis that they can carry bacteria such as Salmonella and E. coli that are harmful to pets and to the people in the household.

The best data on scale comes from an FDA Center for Veterinary Medicine study running from October 2010 to July 2012. In its second year it tested 196 samples of commercially available raw dog and cat food. Fifteen were positive for Salmonella and 32 were positive for Listeria monocytogenes. The FDA concluded that raw pet food was more likely to be contaminated with disease-causing bacteria than the other pet food types tested.

Worth being precise about what that study is: it sampled commercial raw pet food generally, mostly ground meat products, not raw fish specifically. It tells you the category carries real contamination rates. It does not give a rate for raw fish on its own.

The household angle is often underweighted. A dog can shed Salmonella in its stool without appearing sick, which puts anyone handling the dog, its bowl or its waste at risk. In homes with infants, elderly people or anyone immunocompromised, that is the part most worth thinking about.

Bones in raw fish

Raw fish bones are more flexible than cooked ones and less likely to splinter, which is the usual argument for them. They are still thin, sharp and capable of lodging in the mouth, throat or intestine.

The ASPCA's position on bones generally is that dogs can sustain injury or obstruction of the gastrointestinal tract from eating them, and that this can require surgery. The canned-fish exception is the useful one to know: canning softens fish bones enough that sardines and canned salmon can be fed bone-in safely, and the bones add meaningful calcium. Our bones guide covers the cooked versus raw question across all animals.

Does freezing make raw fish safe?

Partly, and the distinction is worth getting right because it is often overstated.

Freezing is an established method for killing parasites in fish intended to be eaten raw, which is why sushi-grade fish is typically frozen before service. So freezing addresses the parasite that carries salmon poisoning disease. Many veterinary sources advise freezing raw salmonids before feeding for exactly this reason.

What freezing does not do is destroy thiaminase, which is an enzyme rather than an organism, or reliably kill bacteria. Freezing suspends bacterial growth; it does not sterilize. So a frozen-then-thawed raw fish still carries risks two and three.

Cooking is the only step that addresses all three at once. If the goal is removing risk rather than managing it, heat is the answer.

What vets actually say, including where they disagree

This is a live disagreement and it deserves to be presented as one rather than resolved by assertion.

The mainstream regulatory and veterinary position is against raw feeding. The FDA advises that raw pet food diets can be dangerous to both pets and the people around them, citing its own contamination findings. The ASPCA groups raw fish with other raw proteins as a food to avoid. The reasoning is straightforward: the bacterial risk is measurable, the parasite risk is potentially fatal, and cooking eliminates both at no nutritional cost worth the trade.

The raw-feeding position is not that these risks are imaginary. Experienced raw feeders generally accept them and manage them: sourcing from suppliers with testing programs, freezing salmonids before feeding, avoiding raw freshwater fish because of thiaminase, rotating proteins rather than building a diet on one fish, and handling raw food with the same hygiene used for raw chicken in a human kitchen. The arguments made in favor are palatability, dental benefit and perceived digestibility.

Where the evidence currently sits: the risks of raw feeding are well documented and quantified, as in the FDA sampling study. The claimed benefits are largely reported by owners rather than demonstrated in controlled studies. That asymmetry is the honest summary, and it is why the veterinary mainstream lands where it does. It is not the same as saying every raw-fed dog will get sick, because most do not.

What is not in dispute by anyone: raw salmonid fish in the Pacific Northwest is dangerous for dogs, and freezing or cooking it first is not controversial advice.

What to do if your dog ate raw fish

• Raw salmon, trout or other salmonid, in or from the Pacific Northwest: call your vet today. Do not wait for signs. The delay is about a week, and early treatment changes the outcome completely.
• Raw fish elsewhere, from a grocery store, in a healthy dog: usually watchful waiting. Monitor for vomiting, diarrhea, fever or lethargy over the next two weeks.
• Raw tuna: the parasite risk is not the salmonid one, so salmon poisoning does not apply. Treat it as a bacteria and bones question, plus the mercury note in the mercury guide.
• Sushi or ceviche: check what else was in it. Soy sauce is a salt load, and onion, garlic or avocado matter more than the raw fish does. See sushi.
• Any dog that develops fever, swollen glands, bloody diarrhea or sudden weight loss within a month of eating raw fish: treat as urgent and tell the vet about the raw fish specifically.
• Puppies, seniors, pregnant dogs and immunocompromised dogs: lower the threshold for calling. They tolerate bacterial illness less well.
• ASPCA Animal Poison Control (888) 426-4435 and Pet Poison Helpline (855) 764-7661, both 24/7. A consultation fee may apply. See also the first 10 minutes.

Ask your vet about your own dog

This page is general education and not a diagnosis or a feeding plan. Whether raw feeding is reasonable for your dog depends on your dog's age and health, your region, your sourcing, and who else lives in your home. Your veterinarian can weigh those together in a way a general article cannot, and is also the person to call if your dog has already eaten raw fish and you are unsure what it was.