Fish Oil for Dogs: Dose, Evidence and Cost in Australia

Do not start fish oil in a dog that has had pancreatitis, and ask your vet before starting it in any dog on other medication. Fish oil is fat, and the doses trialled for arthritis are large amounts of it. Adelaide vet Dr Andrew Spanner puts the risk plainly: “Dogs prone to pancreatitis should not be fed diets with normal to elevated levels of fats and oils. Therefore, fish oil could cause severe illness or even death in these dogs.” Never use cod liver oil as a substitute — it carries vitamin A and D that fish body oil does not, and the dose you would need for joints is not a dose anyone has established as safe with those vitamins attached.

This site has now gone through green lipped mussel, glucosamine and chondroitin, turmeric and the joint supplement shelf generally, and come back with the same answer each time. Fish oil was supposed to be the one that broke the pattern.

It has the best claim of any of them. Dr Andrew Spanner at Walkerville Vet, who has been quoted on this site before, writes: “Fish oil is the most evidence-based dietary supplement for canine arthritis, However, when given on its own it can be disappointing. It is most effective when integrated into a diet.” A 2012 systematic review of nutraceuticals concluded that “the strength of evidence was low for all nutraceuticals except for omega-3 fatty acid in dogs”, and its formal conclusion puts the exception more precisely still: “The evidence of efficacy of nutraceuticals is poor, with the exception of diets supplemented with omega-3 fatty acids in dogs.” Diets, note. That one word is most of this page. A 2022 meta-analysis put omega-3 at the top of its list.

So we went and read the trials. What follows is what they actually say, including the parts that are quoted everywhere and the sentences sitting next to them that are quoted nowhere. Fish oil comes out of this better than anything else on the shelf. It also comes out smaller than you have been told, and most of the evidence did not test fish oil at all.

The three things that make the most difference

Know the difference between “omega-3” and “EPA and DHA”. They are not the same number, the gap between them is enormous, and almost every label in Australia exploits it.
Work out what a product would have to deliver before you buy it. The dose used in the veterinary nutrition literature for osteoarthritis is body weight to the power of 0.75, times 310 mg of EPA plus DHA — about 1,745 mg a day for a 10 kg dog. Most products cannot get near it at their own label dose. Take the number to your vet.
Treat it as one part of a plan. Spanner again: “whichever way you use it fish oil can be helpful, but it’s very unlikely to do enough on its own.” Weight and pain relief do more.

The famous number, and the sentence three lines above it

If you have read anything about fish oil for dogs, you have met this claim: dogs on the fish oil diet were seven times as likely to improve. It comes from a 2010 study by Roush and colleagues in the Journal of the American Veterinary Medical Association, and it is real. Thirty-eight dogs, randomised, double-blinded, measured on a force plate at two university clinics.

Here is the paragraph it comes from, in order, exactly as printed.

“Analysis of force plate results was performed with only data for the most severely affected limb for each dog. None of the values were significantly different between control-food and test-food groups at study start (day 0) or study end (day 90). However, the mean change in peak vertical force for the most severely affected limb between day 90 and 0 was significant for the test-food group (+5.6%; P = 0.01) but not for the control-food group (+0.4%; P = 0.85). Additionally, at day 90, 82% of dogs in the test-food group but only 38% of the dogs in the control-food group had a significant (P = 0.01) improvement in peak vertical force. This corresponds to an odds ratio of 7.0…”

Read those sentences in sequence. The direct comparison between the two groups found nothing. The seven-fold figure comes from counting how many dogs in each group improved, and from the fact that the change was statistically significant within one group and not within the other — which is a different thing from the two groups differing from each other.

The same study’s other outcomes go the same way. Across five signs the vets scored — lameness, weight bearing, range of motion, reluctance to hold up the opposite limb, and pain — not one differed significantly between the groups. And on the owners: “There were no significant differences for any of the scores at day 45 or 90 (data not shown).”

This is now the fourth time this site has run into the same shape. A supplement produces a headline number, and the number turns out to be a within-group change, a responder count, or an owner’s opinion, while the instrument that cannot be flattered finds nothing between the groups.

Most of the evidence did not test fish oil

Of the five trials that built the case for omega-3 in canine arthritis, four tested Hill’s Prescription Diet Canine j/d. In three of those four, the thing it was tested against was a supermarket food, named in the footnotes as “Purina Dog Chow” and “Pedigree Choice Cuts”. The fourth, a dose-titration study, compared three Hill’s foods against each other and so escapes this objection entirely. We come back to what it found.

That is a veterinary therapeutic diet against supermarket dog food. The published composition table shows the two foods differing in crude fibre (9.76% against 1.58%), protein (20.21% against 30.73%), calcium, calorie density, added glucosamine and chondroitin, and L-carnitine. The exposure being tested is not fish oil against no fish oil. It is one whole diet against another whole diet.

The authors saw this coming and addressed it: “One limitation of the present study is that not only the fatty acid content but also some other ingredients differed between the 2 foods. Therefore, it is possible that ingredients other than the fish oil omega-3 fatty acids accounted for the detected effects of food.”

There is one more confound in their own table that the papers do not dwell on. In the force plate study, the dogs on the test food lost weight while the control dogs gained it, and the difference between the groups was significant at P = 0.004. Weight loss is itself one of the best-evidenced treatments for canine arthritis, which we have covered on our page about arthritis treatment for dogs in Australia. A trial in which one arm gets slimmer has a second active ingredient in it.

The “3.5% omega-3” diet is a 0.8% EPA and DHA diet

This is the most useful thing on this page, and it comes straight out of the published composition table.

The trial diet is described everywhere, including in its own abstract, as containing “3.5% fish oil omega-3 fatty acids”. Here is what that 3.5% is made of, from the paper’s own analysis of the test food, as a percentage of dry matter:

EPA — 0.41%
DHA — 0.40%
Alpha-linolenic acid (ALA, from flaxseed) — 2.64%
Sum of all omega-3 — 3.47%

So EPA and DHA together make up about 0.8% of the diet, and roughly three-quarters of the headline omega-3 figure is ALA. That matters because ALA is the plant form, and dogs convert very little of it. Spanner is blunt about it: “Linseed and flaxseed oils contain ALA, which is almost unavailable to dogs.” The veterinary nutrition literature says the same in more careful language — that “ALA is not equivalent to EPA and DHA and the total n-6:n-3 ratio by itself does not accurately describe the fatty acid composition of the diet.”

Carry that arithmetic to the shop with you. When an Australian product tells you it contains a large amount of omega-3, or gives you a “total omega-3” figure and nothing else, you cannot work out from that number how much EPA and DHA your dog is getting. It could be almost none.

One thing the diet did do, and it is the part Spanner emphasises, is invert the ratio. The control foods ran at an omega-6 to omega-3 ratio of 24.33 to 1. The test food ran at 0.71 to 1. That is a thirty-fold shift, and it is achieved by taking omega-6 out as well as putting omega-3 in. Hold onto that, because it is the reason the capsule arithmetic later on is contested.

The one trial nobody paid for came out null

In 2012 a Finnish group published a trial whose title begins with a word you rarely see in a paper: “An un-commissioned randomized, placebo-controlled double-blind study to test the effect of deep sea fish oil as a pain reliever for dogs suffering from canine OA.”

Seventy-seven dogs. Fish oil against corn oil, so an oil against an oil rather than a diet against a diet. The highest supplement dose in the literature — about 110 mg of EPA and DHA per kilogram of body weight per day. Convert that to the metabolic scale the dosing tables use and it comes to roughly 196 mg per kg^0.75 for a 10 kg dog, which is below the 310 the arthritis literature calls a therapeutic dose and about a third of what the top arm of the dose-titration study delivered. Worth holding onto when the result comes in. Force plate measurements plus the validated Helsinki Chronic Pain Index. Sixteen weeks.

The result: “When comparing the two test groups at the end of the trial (16 weeks) there was no significant difference in any of the main outcome variables but owners of dogs that had taken fish oil were significantly happier with the treatment at the end visit and did significantly better at guessing what group their dogs had been in, compared to the placebo group.”

Of the owners whose dogs got the fish oil, 60.6% guessed they were in the treatment group, against 31.4% of the placebo owners who guessed the same. Which brings up a problem that runs through every owner-reported result in this field, and which Spanner names in one line: “I have one concern though: fish oil is very hard to hide due to its smell!”

Now read the same paper’s conclusion to the end, because stopping early would misrepresent it in either direction: “When compared to placebo, there was not a major statistically significant benefit in using deep sea fish oil as a pain reliever in our study population of dogs suffering from osteoarthritis. However, the fish oil treated patients improved significantly in many of the variables, when comparing baseline values to the study-end values within the group, indicating a true but small relief in symptoms. Deep sea fish oil supplementation could be considered a part of the multimodal pain relieving approach…”

A null primary result, and authors who still think it is worth using. Both halves are the finding.

And a second oil-against-placebo trial that went the other way

If the trial above were the whole of the oil-against-placebo evidence, this page would be making a stronger claim than the literature supports. It is not. Mehler and colleagues, in 2016, randomised dogs with osteoarthritis to a fish oil supplement or a placebo and reported: “Seventy-four dogs completed the trial. All clinical outcomes for measuring discomfort, lameness, and joint severity at day 84 and all blood metrics at day 42 and day 84 significantly (p<0.05) improved compared with placebo.” Every clinical outcome, against placebo, at an average dose of about 69 mg of EPA and DHA per kilogram of body weight — well under the dose the null trial used.

Three things about it, every one of which we would have insisted on if the result had gone the other way. There was no force plate and no validated composite pain score, and the authors name that gap themselves: “The use of a refined visual analogue and scoring scale, a validated composite score, and force plate analysis may be helpful in future studies.” So the outcomes are the owner-rated and vet-rated kind that the smell problem above is about. And the funding line reads “The research was partially funded by Veterinarian Recommended Solutions”, with one author disclosed as the owner of the laboratory that ran the blood analyses.

Two supplement trials, opposite results, and the one with the objective measurement is the one that came out null. That is the honest state of it.

The pattern across the literature is worth stating outright. The more objective the measurement and the more the analysis compares the two groups against each other, the smaller the effect becomes. A 2013 force plate trial of a Purina therapeutic diet found its between-group difference at P = 0.211 — though in fairness the authors put that change at an effect size of 0.5, and concluded that the diet “improved the locomotor disability and the performance in activities of daily living” and that “such nutritional approach appears interesting for the management of OA.” The dose-titration study found nothing at twice the baseline dose; at three times it, two of five clinical signs improved significantly — lameness and weight bearing — along with overall arthritic condition and progression of arthritis, those last two at the smallest P values in the whole set. Its authors still called the improvements “relatively modest” in their own discussion.

The meta-analysis, held to the same standard we held it to before

The strongest single citation for fish oil is a 2022 systematic review and meta-analysis that put omega-3 at the top and chondroitin-glucosamine at the bottom. Its own conclusion is unambiguous, and we will print it in full before we go anywhere near it: “Our rigorous approach to meta-analysis allowed us to conclude with certainty that the use of omega-3 products beneficially modulates the painful condition of OA dogs and cats, while the intake of chondroitin-glucosamine has no analgesic effect.”

We leaned on the second half of that sentence on our page about glucosamine and chondroitin. We argued there that the method was capable of finding an effect where one existed, and the omega-3 result was our proof of it. So we owe readers the caveats we did not print then. Here they are. Every one of them works against our own earlier argument.

Two things you need to know about it.

First, its “omega-3-based nutraceuticals” category — the one with the largest effect size — is built mostly from green lipped mussel studies, not fish oil studies. Its own table lists green lipped mussel against six references in that category and fish oil against three. We have written at length about green lipped mussel, and the evidence there is a good deal messier than a top ranking implies. That page also carries the dose by bodyweight for every Australian brand, which runs into the same trouble this page does over what a milligram on a label actually refers to.

Second, and more important, the effect sizes are not what they look like. The review scored each trial on a three-point label — analgesic effect, improvement, or non-effect — “by consensus of the three assessors”, and then computed Cohen’s d across those labels. So a figure like d = 1.19 is a statistic about how three reviewers categorised trials. It does not mean fish oil improved lameness by 1.19 standard deviations in a dog.

The review is candid about the underlying problem: “The one difficulty encountered by previous reviews, and the present one, is the lack of objective data available in studies of these products.” And it makes a point about therapeutic diet trials that deserves to be better known, which we quote in full because its first half is the more deflating one: “the rate of negative responders (using objective kinetic podobarometric assessment) to the introduction of a therapeutic diet is three times higher, while, conversely, the rate of positive responders to a placebo-control diet is up to two times higher.” A dog switched to any carefully formulated food is being given something. Finally, the review’s own conflict-of-interest statement lists work for Nestlé Purina, Royal Canin, Elanco and Zoetis among others, “many of which have led to publications that are included in this systematic review”.

One more, and it is checkable. A peer-reviewed 2024 review of non-drug treatments for canine arthritis describes both of the force plate trials — the Finnish supplement trial and the 2013 Purina diet trial — as having found improvements in peak vertical force. Neither of them found a significant between-group difference in peak vertical force. A within-group change has been read as a treatment effect in a 2024 journal article, which is how a claim like this survives for fifteen years.

The dose, and the arithmetic no label does for you

The dosing figures used across veterinary nutrition come from a 2011 review in JAVMA by Bauer, and they are expressed per kilogram of metabolic body weight — body weight to the power of 0.75.

Osteoarthritis — 310 mg of EPA plus DHA per kg^0.75 per day
Kidney disease — 140 mg. Cardiovascular — 115 mg. Skin and immune conditions — 125 mg
National Research Council recommended allowance — 30 mg
National Research Council safe upper limit — 370 mg

Two things jump out. The arthritis dose is more than ten times the ordinary nutritional allowance. And it sits at 84% of the safe upper limit, which leaves very little room above it — though Bauer’s own table footnotes that the arthritis figure “may be increased (depending on the severity and chronicity of the disorder) up to the NRC safe upper limit”.

One qualification on the 310, in fairness to every product that falls short of it. A 2025 review of EPA and DHA dosing across canine conditions judged the doses shown to be effective in osteoarthritis at 48 to 100 mg of EPA and 20 to 32 mg of DHA per kilogram of body weight — which on the metabolic scale used above is roughly 120 to 235 mg per kg^0.75 for a 10 kg dog, comfortably under Bauer’s figure. So 310 is a widely used target. Nothing in the literature says a dog below it gets nothing, and the arithmetic that follows should be read as what it takes to reach the highest commonly cited dose.

In practical numbers: a 10 kg dog needs about 1,745 mg of EPA plus DHA a day, and a 30 kg dog about 3,975 mg. Spanner publishes the same calculation for Australian owners, using the ceiling figure rather than the target: “raise your dog’s body weight to the power of 0.75 and multiply by 370. As a rough idea, for a 20kg dog the answer would be 3500mg.”

What that means at an Australian chemist

Here is the number that surprises everyone. A “1000 mg fish oil” capsule is a 300 mg capsule. Blackmores Odourless Fish Oil 1000 mg states on its own label: “Natural fish oil 1 g (1000 mg) containing omega-3 marine triglycerides 300 mg as Eicosapentaenoic acid (EPA) 180 mg, Docosahexaenoic acid (DHA) 120 mg.” The gram is the fish oil. The 300 mg is the part that does the work.

So, at 300 mg a capsule:

10 kg dog — about 6 capsules a day
30 kg dog — about 13 capsules a day

Higher strength products cut that down. Swisse Ultiboost Super Strength 2000 mg and Bioglan Double Strength both state 360 mg EPA and 240 mg DHA, so 600 mg a capsule, which halves the count to about 3 and 7. Swisse Ultiboost Odourless High Strength Wild Fish Oil 1500 mg states 270 mg EPA and 180 mg DHA, so 450 mg. In every case the number that matters is that split. The number on the front of the box will mislead you, and where a label gives only a marine triglycerides total with no split, as the section above explains, you cannot get there from a total.

For scale, the same Blackmores label tells a human adult to take two capsules a day for heart, skin and general health, and three for brain and eye health. A 10 kg dog needs two to three times a grown adult’s dose.

And six capsules is also six grams of added fat, sitting on top of whatever the dog already eats. Thirteen capsules is thirteen grams. That is the mechanism behind the pancreatitis warning at the top of this page, and it is why the veterinary review on adverse effects points out that a teaspoon of oil is 40 to 45 kcal. Their worked example uses the cancer dose — 12 to 20 capsules a day for a 10 kg dog, which is 108 to 180 kcal against a resting energy requirement under 400, so a quarter to nearly half of the dog’s daily energy out of one supplement. The arthritis dose is smaller than that. The same arithmetic still applies to it. And in fairness to the same review, its overall verdict is that “the benefits of fish oil intake likely outweigh the risks of toxin exposure and nutrient excess.”

The Australian argument about whether to do this at all

Having done that arithmetic, here is the Australian vet who says the arithmetic answers the wrong question.

Spanner, under the heading “Can You Use Human Fish Oil Capsules?”: “A common question is whether you can use pharmacy fish oil to treat dogs. This is true for skin disease, but false for arthritis. The reason for this is that the food manufacturers, through mysterious means, reduce the levels of omega 6 before adding omega 3.” And: “it would be nearly impossible to achieve either a sufficient dose or an acceptable 6:3 ratio by adding fish oil to an existing diet. Such a dose would probably also be toxic.”

One update to the first half of that. We have since gone through the skin evidence on our page about the itchy dog, and the international guidelines are cooler on omega-3 for skin than “true for skin disease” implies: they grade it at their lowest level of confidence, and say it is no use at all for an acute flare.

His point is the ratio, and the composition table above supports him. The trial diet went from 24 to 1 down to 0.71 to 1 partly by removing omega-6. Pouring oil onto a food that is already high in omega-6 raises the omega-3 and leaves the omega-6 exactly where it was.

And then, a few paragraphs later on the same page, he gives Australian owners the capsule calculator anyway, because the diet is expensive: “The main drawback of j/d is its price. If you find it too costly, then here is a quick formula for the maximum fish oil dose you can give.” He finishes that section with the honest part: “Bear in mind that this will still fall short of ideal doses but it should do something.” His own worked example comes out at eleven grams of fish oil for a 20 kg dog, which is eleven capsules.

That is not a contradiction, and it is the most useful position anyone has put in writing on this. The diet is the instrument the evidence actually tested. The capsules are the affordable fallback, they will not reproduce the ratio, and they should do something. An owner who understands that is better placed than one who has been told fish oil works.

Worth noting that he declares his own interest without being asked: “And a final disclaimer: I sell Hills j/d so I can easily be accused of bias.”

What Australian pet products actually tell you

We went through the pet fish oil products on sale in Australia looking for one number: how many milligrams of EPA and DHA per dose. Most of them do not say.

PAW Fish Oil 500 Veterinary Strength (formerly PAW by Blackmores, now Pure Animal Wellbeing) states 500 mg of EPA and DHA per mL and doses by metabolic body weight, with its own brochure listing osteoarthritis at 310 mg/kg^0.75 and seven pumps a day for a 10 kg dog. That works out at 1,750 mg, which is the therapeutic target almost exactly. It is the only Australian pet product we found whose own label dose reaches it. The brand’s current website no longer lists it, and when we checked on 20 September 2026 the third-party listings we could find were either out of stock — Pet Pharmacy Australia lists it at $35.50 and marks it unavailable — or had been removed from the retailer’s catalogue altogether. Treat it as probably gone.
PAW Dermega Omega Oil describes itself as “formulated with a high concentration of omega-3 and 6 fatty acids” and gives no milligram figure of any kind.
Antinol Plus contains 30 mg of green lipped mussel lipid and 20 mg of krill oil per capsule. It is not an EPA and DHA product and should not be compared on that basis, though it sits on the same shelf.
A powdered skin and coat product states DHA and EPA separately per scoop, and its recommended dose for a small dog delivers under a tenth of the arthritis target.
A widely sold Australian pet omega oil states about 182 mg of EPA plus DHA per 5 mL teaspoon and recommends one teaspoon for a small dog. Reaching the arthritis dose would take roughly ten teaspoons a day, and the product is cod liver oil based, which makes that volume a bad idea for the reasons at the top of this page.

The pattern is uncomfortable. The product that stated a usable figure and dosed to metabolic weight is the one that has quietly left the brand’s own catalogue, and the products that replaced it on the shelf state less.

We could not check the APVMA register for these products: the portal was down for maintenance throughout 20 September 2026. What we can say is that the APVMA’s own rule is that a complementary animal health product may “make only general health claims, not specific therapeutic claims that refer to the prevention (or reduction), cure or alleviation of a specific disease or condition.” Several Australian joint products name arthritis outright in their marketing, and PAW’s own Dermega omega oil page lists “REDUCES INFLAMMATION” among its key benefits, which is a pharmacological claim rather than a general health claim.

The Australian heat problem, and the argument about it

Fish oil oxidises. It is what rancid means, and it is accelerated by warmth, light and time.

In 2019 a group published “Oxidation of fish oil supplements in Australia” in the International Journal of Food Sciences and Nutrition, testing 26 supplements bought on the Australian market. Their result: “38% exceeded the limit for primary oxidation, 25% exceeded the limit for secondary oxidation and 33% exceeded the limit for total oxidation, according to international recommendations.”

The same paper found something Australian shoppers should know: “Four specially marketed supplements were found to deliver significantly lower amounts of fish oil per capsule (165 vs. 577 mg, p = .007), yet cost significantly more on a per gram basis ($2.97 vs $0.39, p < .001).” And there was no difference in oxidation between the premium products and the ordinary ones.

A 2015 New Zealand study of 32 products found worse: 83% above recommended peroxide levels and most containing under two-thirds of their labelled EPA and DHA.

Now the other side, because it exists and it matters. A 2016 paper tested ten Australian and New Zealand products, bought in a single Melbourne suburb, and found all ten met the recommended peroxide value, with EPA and DHA averaging 109% of label claim. The row quoted less often is the second oxidation test: “Anisidine value (pAV) met the international recommended level for eight of the 10 products, with two products known to contain flavorings that interfere with the pAV test.” Two of ten did not pass it. That study’s analyses were funded by the Australian and New Zealand Omega-3 Centre, an industry body, and one of its authors was affiliated with an omega-3 ingredient supplier. None of that makes it wrong. It does mean it is not independent. So the pattern is that independent testing has repeatedly found problems and industry-linked testing has repeatedly not.

The pet products have been tested once, and not for oxidation. A 2025 analysis of canine fish oil supplements examined label accuracy and contaminants across products sampled in 2010 and again in 2021, and found that “Most product labels (21/22) included minimum content claims for eicosapentaenoic acid and docosahexaenoic acid concentrations, and six of those products did not meet the minimum claim for one or both fatty acids.” Six labels out of twenty-two overstated what was in the bottle. PCBs and most heavy metals were not detected, which is the reassuring half. Vitamins A and D were detected in some products that did not declare them, which is not — and which is the cod liver oil problem at the top of this page arriving inside a product that does not call itself cod liver oil.

The storage angle is where Australia is genuinely different, and we will separate what is sourced from what is ours. The Australian study notes that “storage in darkness and at low temperatures is recommended for bulk fish oil before and during manufacture, yet no such recommendation exists for finished encapsulated fish oil products.” Australian labels reflect that: Blackmores capsules say “Storage Temperature: Below 30°C”, Bioglan says “Store below 25°C”, and several brands say nothing at all. The liquids are stricter — one Australian liquid fish oil directs “Store below 25°C, in a cool dry place out of direct sunlight. After opening, store at 2°C to 8°C (Refrigerate, do not freeze)”, and an Australian pet omega oil says to refrigerate after opening.

Our own observation, offered as that: a cupboard on an external wall in a Queensland February is not reliably below 30°C, and a car is not below it for a minute. We found no Australian manufacturer, vet or regulator addressing summer ambient temperature in relation to fish oil. Keep the bottle in the fridge, buy smaller bottles more often, and throw it out if it smells sharp or sour.

What fish oil is and is not evidence-based for

Beyond joints, the picture varies a great deal and the marketing does not.

Skin and itch — the best supported of the non-joint uses, and the one Spanner says pharmacy capsules genuinely do suit. Still modest — and in one trial, in the reviewers’ own words, “no association was detected between the total fatty acid intake or omega-6–to–omega-3 ratio and clinical scores.”
Kidney disease — there is canine data, but the key study used a surgical model and a dose the reviewers themselves describe as more than twice the safe upper limit. That is not a dose to copy.
Heart disease — dogs in heart failure have low blood EPA, and supplementation is described in the literature as something that “may” help. Mechanism, not proof.
Cognition — the trial everyone cites used a diet enriched with DHA and antioxidants and mitochondrial cofactors, and the authors say so: “the extent to which DHA played a role in the improvements remains to be established.”
Cancer — there is one canine finding. In a trial of dogs with lymphoma the reviewers report that “increasing the serum DHA content in the dogs was associated with longer disease-free intervals and survival times.” That is an association with a blood level inside a small trial. It falls a long way short of showing that feeding fish oil extends a dog’s life, and the reviewers still file the area under emerging.
Gut disease — a 2025 review of EPA and DHA dosing concluded that gastrointestinal disease “does not appear to benefit from omega-3 supplementation”.

That last review has to be read past the line we just took from it, because the rest of it cuts against the tone of this list. Its authors conclude: “In conclusion, the reviewed EPA and DHA doses appear effective for atopic dermatitis, osteoarthritis, cardiac disease, hyperlipidaemia, and cognitive and behavioural disorders. Further research is needed to clarify efficacy in gastrointestinal and oncological conditions.” Heart disease and cognition sit in their effective column. We have set out above why we read the underlying trials more cautiously than they do — the heart evidence is largely mechanistic and the cognition trial changed four things at once — but a reader is entitled to both readings, so here are both.

What it costs in Australia

Prices checked 20 September 2026 at the retailers named. The costs below are what it takes to reach the arthritis dose, which is a different thing from what it costs to follow a product’s own label.

Swisse Ultiboost High Strength 1500 mg, 400 capsules at $39.99 from Chemist Warehouse — about $12 a month for a 10 kg dog, about $26 for a 30 kg dog.
Blackmores Odourless 1000 mg, 400 capsules at $38.49 — about $17 a month for a 10 kg dog, about $38 for a 30 kg dog.
PAW Fish Oil 500, 200 mL at a listed $35.50 — about $19 a month for a 10 kg dog, about $42 for a 30 kg dog, if you can find it in stock, which on 20 September 2026 we could not.
Hill’s j/d — a 12.5 kg bag runs roughly $178 to $262 depending on retailer and membership. Feeding it to a 10 kg dog is in the order of $80 to $95 a month, and to a 30 kg dog closer to $200. That last figure is our estimate from the published energy density and price, not a figure from Hill’s own feeding guide, so treat it as indicative.
The pet-specific powders and oils, taken to the arthritis dose rather than their label dose, come out at $80 to $200 a month — five times the cost of chemist capsules for the same amount of EPA and DHA.

Cost, in other words, is not the obstacle to getting EPA and DHA into a dog. The obstacles are the number of capsules, the fat that comes with them, the ratio problem, and the fact that the diet with the actual trial evidence behind it is the expensive option.

This site earns nothing from any product named on this page, and there are no affiliate links on it.

The five things not to do

Do not buy on the “omega-3” number

If a label gives a total omega-3 figure, a “marine triglycerides” figure, or a fish oil weight without splitting out EPA and DHA, you cannot calculate a dose from it. Put it back.

Do not use cod liver oil

It is a different product with vitamin A and D in it. The joint doses on this page assume fish body oil and nothing else.

Do not go above the ceiling

370 mg per kg^0.75 per day of EPA plus DHA is the National Research Council’s safe upper limit, and the arthritis dose is already at 84% of it. More is not better here, and the adverse effects reviewed in the veterinary literature include gastrointestinal upset, altered platelet function, reduced vitamin E, and the calories.

Do not start it in a dog with a pancreatitis history without asking

Covered at the top of this page and worth repeating, because the risk is not theoretical and the dose is large.

Do not treat it as the plan

Weight control has better evidence than any supplement on the market. Pain relief that a vet prescribes has better evidence again. Fish oil belongs alongside those, and our page on the older dog that has slowed down covers what else should be on the list before you reach for a bottle.

Fish oil is the cheapest thing on the joint shelf by a wide margin, which is easier to see set against the rest. We have costed eight Australian joint supplements for dogs per day at their label dose.

The short version

Fish oil has the best evidence of any joint supplement for dogs. That is a statement about how weak the rest of the field is.
The famous “seven times as likely to improve” sits three sentences below “None of the values were significantly different between control-food and test-food groups.” Both are in the same paper.
Three of the five headline trials tested Hill’s j/d against supermarket food — named in the footnotes as Purina Dog Chow and Pedigree Choice Cuts — two whole diets differing in fibre, protein, calcium, glucosamine and calories — and in one trial the test dogs lost weight while the controls gained, P = 0.004.
The “3.5% omega-3” diet is 0.41% EPA and 0.40% DHA. Most of the rest is flaxseed ALA, which dogs barely convert.
The one trial nobody commissioned found no significant difference in any main outcome — and its owners could guess which group they were in, because fish oil smells.
A second oil-against-placebo trial found every clinical outcome improved against placebo. It had no force plate, no validated pain score, and part funding from the supplement’s maker. Both trials belong on the page.
The 2022 meta-analysis’s top-ranked “omega-3 supplement” category is mostly green lipped mussel, and its effect sizes are computed on three reviewers’ labels rather than on dogs.
The dose used in the literature is body weight to the power of 0.75, times 310 — about 1,745 mg of EPA plus DHA a day for a 10 kg dog, against a safe ceiling of 370 on the same scale.
A “1000 mg” capsule holds 300 mg of EPA and DHA. That is six capsules a day for a 10 kg dog and thirteen for a 30 kg dog, and six to thirteen grams of added fat.
An Adelaide vet argues the capsules are the wrong instrument, because you cannot fix an omega-6 to omega-3 ratio by pouring oil on top — and then publishes the capsule calculator anyway, because the diet is expensive, with the words “it should do something”.
Independent Australian testing found a third of pharmacy fish oils over the oxidation limits, premium products no better than cheap ones — and the one analysis of pet fish oils found six of twenty-two labels overstating their own EPA or DHA.

Sources

This page is general information, not veterinary advice, and it cannot tell you whether fish oil is right for your dog or at what dose. That calculation needs your dog’s weight, its other conditions and its other medications, and it belongs to a veterinarian who has examined it. Do not start a high-dose oil in a dog with a history of pancreatitis without veterinary advice.

Fish Oil for Dogs: The Evidence, the Dose, and What It Costs in Australia