Oils
Every comedogenic list prints one number per oil. Fulton, whose table those numbers come from, printed two things the lists never carry: an asterisk on the section heading meaning "results depend on source of raw material", and, for six oils, a second number in parentheses for the refined version of the same oil. Sesame oil scored 3, and a refined grade of it scored 1. None of the dozen lists we checked print the 1.
What they are
Triglycerides, nearly all of them: three fatty acids esterified to a glycerol backbone, pressed or extracted from a seed, a nut, a fruit or an animal. What distinguishes one from another is largely the fatty-acid profile — how long the chains are and how saturated. That profile is one of the things the comedogenicity literature reached for, and we should not pretend it is the only one: Fulton's own results turn as much on source, refining, residual lipids and the vehicle an ingredient was tested in as on bulk fatty-acid percentages. ONE member of this family is not a triglyceride, and it is worth knowing which. Jojoba is a liquid WAX: its fatty acids are esterified to long-chain alcohols rather than to glycerol, which is a different molecule doing a different job. It sits in the oils aisle and gets scored on the same scale as sunflower oil, and it is not the same kind of substance. One thing this page used to get wrong, because it explains the section below. We told you MINERAL OIL was a member of this family. It is not: we file mineral oil, squalane and petrolatum as HYDROCARBONS, in their own family, because they are not fats at all — and the table below will not show you any of them. FULTON MADE THE SAME MISTAKE, and that is the useful part. His section "VI. Oils*" holds cocoa butter, coconut butter, squalane and mineral oil alongside the sesame and the corn oil. It is a FORMULATOR'S category — things that behave like an oil phase — not a chemical one. So when we say the asterisk on that heading governs everything under it, we mean everything: the butters and the hydrocarbons too, none of which are in this family as we file it. His section and our family are different sets.
What the evidence actually says
Here is the thing this page exists to say, and it is not a matter of interpretation. It is printed in the paper. Fulton's 1989 table is organised into numbered sections. Section VI is headed, in full: "VI. Oils*". That asterisk is defined once, at the foot of the table, and it means: "Results depend on source of raw material." It is not attached to one row. It is attached to the HEADING, which means it governs every oil and every butter Fulton scored. Cocoa butter's famous 4 is under it. Coconut butter's 4 is under it. Olive oil, castor oil, mineral oil, squalane, all of them. We did not notice this for weeks, because we transcribed the rows and not the headings, and an external fact-check on cocoa butter caught us. So we are not going to be smug about it. But the industry is quoting a table while deleting the sentence that tells you what the table means. And there is a second notation, which is stranger and more useful. Six of Fulton's oils carry TWO numbers, the second in parentheses. Almond oil 2(1). Apricot kernel oil 2(1). Olive oil 2(1). Avocado oil 3(2). Mink oil 3(2). And sesame oil, 3(1). The parenthetical is defined by a second footnote: "Parentheses indicate results using 'refined' oil." Fulton says the same thing in his prose, plainly, without a hint that he considers it a footnote: "Among the natural oils such as sesame oil, avocado oil, and mink oil, the results are improved when a more refined oil is used." Sit with the sesame number for a second. The oil he tested scored 3, which Fulton's own key calls borderline. A REFINED sample of it scored 1, which his key calls "not considered significant." That is not a rounding difference. It is the gap between an ingredient a list tells you to avoid and an ingredient the same paper calls unremarkable, and it is the same oil off a different production line. One precision, because it is the kind we go after other people for. Fulton does not print the word "unrefined" anywhere. He prints a number, and then a refined number in brackets. The first sample is the comparison grade, and we do not know what it was — virgin, cold-pressed, solvent-extracted, or simply whatever arrived. An earlier draft of this page called it "unrefined sesame oil" as though that were on the page. It is not. What the table establishes is that two grades of the same oil gave two different answers. We checked a dozen of the lists that publish these numbers. Not one of them printed the refined score, and not one carried the asterisk. We are not going to tell you that NO list on the internet does, because we have not read every list on the internet, and a claim we cannot check is the thing this page exists to object to. So: for an oil, the comedogenic number is not a property of the plant. It is a property of the material that arrived in the jar, and Fulton said so twice in one table — once in the asterisk, once in the parentheses. What he does NOT tell us is which step does the work: he reports no pressing method, no solvent, no oxidation state, no free-fatty-acid content, and never says which part of refining changed the result. So we can say the score tracks the source and its processing. We cannot hand you a mechanism, and we previously blamed "the press", which is a detail he never recorded. He was blunter still about one of the section's non-fats, the one everybody argues about: "Mineral oil presents a complex problem: some sources are acceptable; others are not." A label does not tell you which source. Neither can we. That is the honest state of this entire family, and it is why almost every oil we rate sits at Caution rather than at a confident number.
What the evidence doesn’t tell you
This does not mean oils are fine. It means the number is doing less work than it appears to, in both directions, and we are not going to swap one overconfident story for another. The refined figure is still a rabbit's ear. Everything above comes from the same 1989 assay we discount everywhere else on this site: an ingredient diluted to roughly 10% in propylene glycol, applied to the inner ear of three New Zealand albino rabbits, five days a week for two weeks, follicles scored for keratosis. Draelos and DiNardo's 2006 re-evaluation tested FINISHED PRODUCTS on small human panels and concluded that isolated rabbit-ear ingredient scores are a poor guide to what a product does on a human — it was not a same-material human re-test of any of these six oils, and we previously implied a generality it does not have. The 2025 JAAD Reviews survey found that no standardised comedogenicity test exists at all. Those objections apply to the reassuring number in the parentheses exactly as hard as to the alarming one outside it. We cannot tell you what is in your bottle. "Sesame Oil" on an INCI list does not say refined or unrefined, cold-pressed or solvent-extracted, and on the labels we have looked at — EU and US cosmetic ingredient declarations — nothing requires it to. So the practical value of Fulton's parenthetical is limited: it tells you the score is unstable, not which end of the range you are holding. That is genuinely useful and genuinely incomplete, and we would rather say so than pretend the asterisk resolves anything. The papers also disagree, and the almond case is the one to look at — including for what WE did with it. Morris and Kwan scored SWEET ALMOND OIL 3, and that is the number we carry, because we have read their table. Fulton prints "Almond oil 2(1)" — and that is all he prints. No species, no supplier, no sweet-versus-bitter, no grade. An earlier draft of this page called that "two laboratories, one material." We do not know that it is one material. It probably is: sweet almond oil is what the cosmetic industry means by almond oil. But PROBABLY is not a standard we accept from anybody else, and quietly turning one material name into another is the failure this site is about. We did it inside the paragraph accusing the literature of doing it. What is documented: Morris and Kwan tested Sweet Almond Oil at 3; Fulton tested an unspecified Almond oil at 2, and a refined grade of it at 1. Whether those are the same substance is not established. Now the claim that was simply false, and our own coconut oil page disproves it. This page said "nobody has counted comedones on a human face after using any of these oils" and "the whole family rests on rabbits." Both are wrong. Ungaksornpairote and colleagues (2015) applied virgin coconut oil to the WHOLE FACES of 30 women, twice daily for a month, and thirteen of the twenty-six who started acne-free developed acne — a family member, a human face, and a finding we cite at length two clicks away. Gala and colleagues (2021) applied coconut oil to human backs as a positive control and graded comedones on a 0–3 scale. So the honest criticism is not that human evidence does not exist. It is that it is thin, uneven and concentrated in one or two oils: coconut has direct human facial exposure; most of the rest of this family has no human comedone data we can find at all; the studies that exist differ in design, rarely isolate the oil, and almost never count comedones on a face. The numbers you will meet for the other twenty oils remain rabbit-derived. That is the true statement, and it is nearly as damning as the false one — which is exactly why there was no excuse for reaching for the false one.
All 22 oils we rate
The legacy score is printed exactly as it appears in the paper, including the ranges, the asterisks and the parentheses. Those marks are not decoration. They are the paper telling you what its own number is worth.
Sources
- Fulton JE. "Comedogenicity and irritancy of commonly used ingredients in skin care products." J Soc Cosmet Chem 40:321–333 (1989) The primary source for essentially every comedogenic number in circulation, read off the journal scan rather than via a list quoting it. Three things on this page come out of it and out of nothing else: the section heading "VI. Oils*"; the definition of that asterisk, "Results depend on source of raw material"; and the parenthetical second scores for the six oils where a refined grade was tested. The lists that republish his numbers carry none of the three.
- Draelos ZD, DiNardo JC. "A re-evaluation of the comedogenicity concept." J Am Acad Dermatol 54:507–512 (2006) The published objection to the assay all of the above rests on. It tested FINISHED PRODUCTS on small human upper-back panels and found that an ingredient tested alone says little about the product it ends up in. Cited because it cuts against the reassuring bracketed numbers exactly as hard as against the alarming ones outside the brackets, and a page that used it in one direction only would be choosing its evidence. It is NOT a human re-test of these oils, and we previously described it as though it were.
- Morris WE, Kwan SC. "Use of the rabbit ear model in evaluating the comedogenic potential of cosmetic ingredients." J Soc Cosmet Chem 34:215–225 (1983) An INDEPENDENT rabbit-ear assay, six years before Fulton's famous table and on the same 0–5 scale. It is not a republisher of him, and it disagrees with him in print. Cited for the awkward part: it scores SWEET ALMOND OIL 3, where Fulton prints "Almond oil 2(1)" and specifies no species, supplier or grade. We previously called that "two labs, one oil". We do not know it is one oil, and asserting it would be the exact silent normalisation of one material name into another that this entire site was built to catch. Two labs, two names that are probably the same cosmetic material, and three numbers.
- Ungaksornpairote C et al. "A Pilot Study of the Effects of Virgin Coconut Oil on the Skin." Thai J Dermatol 31(3):158–169 (2015) THE PAPER THAT DISPROVES A SENTENCE THIS PAGE USED TO CARRY. We wrote that nobody has counted comedones on a human face after using any of these oils, and that "the whole family rests on rabbits." This is a family member, on thirty whole human faces, twice daily for a month: thirteen of the twenty-six who started acne-free developed acne, and facial sebum rose significantly. It is uncontrolled, unrandomised, all-female, and acne was an adverse-event observation rather than a powered endpoint — which is exactly why we rate coconut oil on it at MODERATE and not High. But it exists, we cite it at length on the coconut oil page, and a universal claim contradicted by our own bibliography two clicks away is not a slip of the pen. It is the failure this site is about.