The label 'non-comedogenic' suggests a product won't clog your pores. That is the implied promise on tens of thousands of moisturizers, sunscreens, and serums. Behind it sits a 1970s test on rabbit ears, no FDA regulation, and a 1982 study by the test's own inventor showing the rabbit results often did not match what happened on human skin. The word is functionally a marketing term wearing a lab coat.
Where the term actually comes from
In 1972, dermatologist Albert Kligman and his colleague Otto Mills published a paper coining the phrase 'acne cosmetica,' arguing that certain cosmetic ingredients could trigger pimples in adults who otherwise had clear skin. To screen ingredients for this property, Kligman adapted an existing assay: applying test substances to the inner ears of albino rabbits for two weeks and counting the microscopic blackheads that formed in the follicles.
The rabbit ear was chosen for a specific reason. Rabbit follicles are unusually large and unusually responsive, so the test produced clear, measurable results in a short timeframe. It was efficient. It became the industry default. The numeric scale you have probably seen on skincare blogs, where ingredients get rated 0 to 5, traces back to that same model. The American Academy of Dermatology endorsed a version of it in 1989. That endorsement is still being cited today.
Kligman ran the test for industrial chemicals first
The rabbit ear assay was not originally designed to evaluate face cream. Kligman developed it to assess industrial chemical exposure in factory workers handling things like cutting oils and tar derivatives, substances known to cause severe occupational acne. The test was useful in that context because the chemicals being studied were aggressive and the question was binary: does this stuff cause skin lesions in workers, yes or no.
When the cosmetics industry adopted the same assay to screen ingredients in finished consumer products, the question changed but the method did not. A test built to flag harsh industrial solvents was now being used to evaluate a 2 percent concentration of jojoba ester in a moisturizer. The signal-to-noise ratio collapsed almost immediately.
Kligman built the rabbit ear test for factory workers handling industrial solvents. Marketing took it from there.
The 1982 paper that should have ended this
Kligman himself spotted the problem. In 1982, he and Mills published a follow-up study using a human model, applying test substances to the upper backs of young men with large follicles for one month and analyzing the follicular contents directly. The result was inconvenient.
Several ingredients that produced strong comedogenic responses on rabbit ears produced little or no comedone formation on human skin. Other ingredients flagged as 'mildly comedogenic' in the rabbit model turned out to be benign on people. The paper concluded plainly that the rabbit assay was overly sensitive and not consistently predictive of acne risk in humans. The rabbit ear, biologically, was reacting to things human skin could handle.
That paper was published 44 years ago. The ratings industry still cites lists derived primarily from the rabbit data.
What 'non-comedogenic' on a label actually means
Legally, in the United States, the term means nothing. The FDA does not define 'non-comedogenic.' There is no required test, no required threshold, no enforcement. A brand can write the word on the bottle without conducting any clinical study at all. Some brands do run human-use studies on their finished formulas. Most just check the ingredient list against a comedogenicity database and call it done. The consumer has no way to tell which type of product they are holding.
The Modernization of Cosmetics Regulation Act, the biggest US cosmetics regulation overhaul since 1938, is rolling out through 2026. It requires safety substantiation, adverse event reporting, and facility registration. It does not define 'non-comedogenic.' That gap is intentional. Defining the term would require the FDA to choose a test method, and the existing methods are either ethically dated, not predictive, or expensive to run at scale.
Why ingredient-level ratings are misleading
The 0 to 5 comedogenicity scale you see on websites and skincare apps is a useful starting point, but it has three structural problems that consumers rarely hear about.
- Concentration is ignored. A coconut oil rating of 4 was generated using essentially undiluted oil. The same ingredient at 1 percent in a humectant-rich serum behaves nothing like the test condition. Most ratings discard concentration entirely.
- Formulation chemistry is ignored. An emulsifier, a film-former, or a fatty alcohol in isolation may rate higher than the same ingredient inside a stable cream. The matrix matters. The rating doesn't capture it.
- Individual variability is ignored. Acne-prone skin is heterogeneous. A 2024 review in the Journal of the American Academy of Dermatology specifically flagged that comedogenicity testing rarely accounts for follicle size differences, and follicle anatomy varies meaningfully across skin tones and individuals.
So when you see an app tell you a product contains a '4-rated' ingredient and conclude you should avoid it, the underlying data is a 1970s rabbit ear test on the pure substance, scaled by a system that the test's inventor partially repudiated, applied to a finished formula it was never designed for.
What a useful claim would actually look like
'Non-comedogenic' on a label is doing a job. It is supposed to signal that the brand has thought about acne-prone users. The signal is just unverifiable in its current form. A more honest version of the same claim, the kind a consumer can actually evaluate, looks like one of the following:
- 'Tested in a 4-week clinical study on 30 acne-prone subjects.' This is a finished-formula human use study. It costs the brand real money. It produces an actual answer for that specific product.
- 'Non-occlusive formulation.' A statement about the chemistry rather than the outcome. Verifiable by reading the ingredient list and looking for known occlusive emollients.
- 'Free of [specific ingredient class] linked to acne in [specific study].' Specific. Citable. Falsifiable.
'Non-comedogenic' alone tells you the brand wants you to think about acne. It does not tell you what they did to earn the claim.
What to actually do with this
You do not have to throw out everything labeled 'non-comedogenic.' The label is not a lie. It is just not evidence. A few practical moves:
- Read the ingredient list. Isopropyl myristate, isopropyl palmitate, and certain high-concentration plant butters are reasonable things to flag if you have aggressively acne-prone skin. Not because the rabbit data is gospel, but because human-use evidence has accumulated for some of them.
- Patch test new products on the side of your jaw for two weeks before committing to full-face use. This is the only test that actually applies to your skin.
- Look for brands that publish their human-use clinical study summaries. Most don't. The ones that do are giving you the information the label cannot.
- Treat the comedogenicity rating on a website as a hypothesis, not a verdict. A 4-rated ingredient at 0.5 percent in a well-formulated emulsion may be a non-issue for you. A 1-rated ingredient at 30 percent might break you out.
At Skinventry, we track every ingredient in every product on your shelf and flag the marketing claims that do not have a regulated definition behind them. 'Non-comedogenic' is one of those claims, and we tell you what kind of testing, if any, the brand actually ran to back it up.