Is your fish oil rancid? Read the test first

"Is my fish oil rancid?" is one of the most searched supplement questions there is, and the answers online tend to quote a single scary percentage. The trouble is that "oxidized" is not something you see directly. It is the output of two lab tests and a bit of arithmetic, each with its own blind spots. Before you trust any percentage, it helps to know what those tests read.

The three numbers

Omega-3 fats (EPA and DHA) have many double bonds, which makes them easy to oxidize. Oxidation happens in stages, and the standard tests each catch a different stage:

Peroxide value (PV) measures the first products, lipid peroxides. In the 2015 study it was done by iodine titration: peroxides release iodine, and you count how much it takes to clear the color.

Anisidine value (p-AV) measures the later breakdown products, mainly aldehydes. The oil is reacted with a reagent called p-anisidine and read on a spectrophotometer at 350 nm. In other words, it is a color reading.

TOTOX is not a third test. It is a formula: 2 × PV + p-AV, meant to capture both stages in one number.

The limits most studies use are the voluntary industry ones from GOED (the Global Organization for EPA and DHA Omega-3s): PV under 5, p-AV under 20, TOTOX under 26. Those are not the only limits. The European and British Pharmacopeias and Australian regulators use looser ones (PV 10, p-AV 30, TOTOX 50, as the 2017 paper summarizes them). And per the 2015 paper, "industry standards are based on palatability, as there are insufficient data to set standards based on health effects." The thresholds describe taste and quality, not a proven line between safe and harmful.

Study one: "highly oxidised"

Benjamin Albert and colleagues at the University of Auckland's Liggins Institute bought every encapsulated fish oil they could find in New Zealand and published the results in Scientific Reports in January 2015. Of 36 products tested for oxidation, 83% exceeded the PV limit, 25% exceeded the p-AV limit, and 50% exceeded TOTOX. Only 3 of 36 passed all three. On content, the 32 products checked for EPA and DHA held on average 68% of what the label claimed, and only 3 of 32 met or beat their label.

Two further findings are useful for any shopper. Best-before date, price, and country of origin were all poor predictors of quality. And the authors made a point that is easy to miss: peroxides form early and then break down, so "a low PV is consistent with both minimal and severe levels of oxidation." A clean PV on its own doesn't clear an oil.

One correction belongs in any honest summary. In November 2016 the journal published a corrigendum: the PV results were in meq/kg, not meq/l as printed, and the anisidine formula in the methods had a typo. The journal describes both as typographical errors.

Study two: "appears overstated"

In May 2017, Gerard Bannenberg and colleagues published a retest in the same journal. They bought 47 encapsulated fish oils from New Zealand online retailers and sent blinded samples to two or three accredited labs each. Their results: 72% met the PV limit, 9% fell below 90% of their EPA and DHA label claim, and among unflavored oils 86% met the p-AV limit and 77% met TOTOX. Against the looser pharmacopeia limits, compliance was higher still. They called the 2015 title's "highly oxidized" conclusion "overstated."

Before weighing that, the disclosure: the study was financed by GOED, two authors work for GOED (a trade association of more than 200 omega-3 companies), and the other five are analytical chemists at labs that are GOED members. That is stated plainly in the paper. It doesn't make their measurements wrong. It does mean the side with a commercial interest in a clean result ran the retest, which is exactly why the arguments have to stand on chemistry rather than reputation.

Where the gap comes from

Read closely, the two papers are not far apart on secondary oxidation (25% of products over the p-AV limit in 2015, 14% of unflavored oils in 2017). The big disagreements are PV and label content. The 2017 paper offers three explanations, and each one is a general lesson about this kind of testing.

1. Handling can create peroxides. The 2015 team pooled oil from 8 to 24 capsules per product and stored the capsules at 4°C in amber jars before testing. The 2017 team argues that oil out of the capsule oxidizes quickly in air, so any delay or the wrong labware can push PV up in the lab rather than on the shelf. Their evidence is a ratio: in the 2015 data, 38% of products had PV higher than p-AV, a pattern they say was rare across other testers' published results and is "suggestive of recent oxidation." That is an inference about the other lab's work, not a demonstration. But it is the right question to ask of any study: was the oil oxidized when it was bought, or after it was opened?

2. Flavors fool the anisidine test. This is the point I find most useful, and it is a measurement point. p-AV is an absorbance reading after a color reaction. Anything in the oil that reacts with the reagent or absorbs at that wavelength gets counted as oxidation. The 2017 paper states that "the validity of p-AV and TOTOX results is limited for many marine oil products that contain flavorings," and that the interference generally inflates the result. In their own samples, 15 of 25 flavored oils read above the p-AV limit, but a separate method built for flavored oils flagged only 4 of the 21 it could test. GOED's own quality standard excludes flavored oils for this reason. So a study that applies p-AV limits to lemon-flavored capsules may be measuring the lemon. (My own research is microwave spectroscopy, not this UV-visible assay, but the rule is the same in every spectroscopy lab: a reading tells you what absorbed, not why.)

3. The pass mark changes the pass rate. On label content, the 2015 paper counted a product short if it held less than 100% of its claim. The 2017 paper used Australian rules, which require at least 90%, and also argued that results must be expressed in the same chemical form as the label. That threshold difference can't explain all of the gap, though: an average of 68% of claim fails either rule. The 2015 paper didn't name its products, so the two lists can't be matched one to one.

What both papers support

Strip away the argument and some common ground remains. Oxidation creeps up over shelf life: when the 2017 team retested five products a year later, past their expiry dates, PV and TOTOX had risen close to the limits (their EPA and DHA content held steady). Some products fail even in the industry-funded data, 28% on PV. Neither paper measured health effects, and the 2015 authors note that the health implications of oxidized fish oil "remain unclear," with mostly animal evidence and one short human study to go on.

For a buyer, that suggests a few practical habits rather than a verdict:

Ask for batch numbers, not marketing. A certificate of analysis that reports PV and p-AV for your batch tells you more than "molecularly distilled" or "pharmaceutical grade" on the label.

Know that flavored oils are hard to test. A high anisidine value on a flavored product may not mean rancid, and a passing one was likely measured some other way. Ask how.

Don't read the date as freshness. In the 2015 data, time to best-before told you nothing about oxidation. Oxidation still rises with age, so storage and turnover matter.

Read "X% of fish oils are rancid" as a lab result, not a fact about your bottle. Which limits, which labs, flavored or not, and how long the oil sat open before the test all move that percentage.

Not medical advice. This is educational analysis, not a recommendation — a study is not a prescription. Talk to a qualified clinician before acting on anything you read here. Full disclaimer →

Disclosure, plainly: I founded and run Shroombiosis (a company I run), which formulates and sells supplements. That is a commercial stake in how supplement research gets read, and the same testing questions apply to anything I make. I have no relationship with any author, lab, trade group or brand discussed here. Nothing here is sponsored and no link earns a commission; here's the full policy. A recommendation with no stake at all: for performance nutrition, Die Tryin Co. is a fellow combat-veteran-owned brand I'm glad to point people to. I don't own it and earn nothing from the link.