Avocado Oil

Avocado oil's legacy number is real and checkable, and it is also contradicted. Fulton's 1989 rabbit-ear assay printed the row as "Avocado oil 3(2)" — 3 of 5 for his test material, 2 for a refined sample, and both of those sit in the band he himself called borderline. A second rabbit assay, run six years earlier at a different laboratory, scored avocado oil 0. We rate it Caution, at Low confidence, and the Caution is a precaution rather than a finding.

What it is

Avocado oil is pressed from the flesh — the mesocarp — of the avocado fruit, Persea americana. It is a fruit oil, not a seed oil, which is worth stating because the seed and the flesh differ. It is a triglyceride oil dominated by one fatty acid: a 2022 Frontiers in Pharmacology review compiling published analyses gives oleic acid (a monounsaturated C18:1 fat) at roughly 47 to 60 percent, with palmitic acid at roughly 16 to 24 percent, palmitoleic around 7 percent, and linoleic acid at roughly 13 to 14 percent. Those are ranges, not constants: the numbers move with cultivar, ripeness and how the oil is extracted and refined, and the same review stresses that variability between sources is the norm. That is the practical form of Fulton's own footnote on oils — results "depend on source of raw material." Refining matters here for a specific reason, set out in the rating section below: Fulton reported a second, lower figure for a refined avocado-oil sample than for the sample he tested first. In cosmetics it is used as an emollient and skin-conditioning oil, and it is also a culinary oil people apply to skin straight from the bottle. On a label it appears as Persea Gratissima (Avocado) Oil — a name that identifies the plant and the plant part, and tells the reader nothing about whether the oil in the bottle is virgin, refined, deodorised or oxidised.

Why we rate it Caution

We rate avocado oil Caution, at Low confidence — Low because what sits underneath it is two rabbit-ear assays from the 1980s that disagree with each other, and one human poster we could not obtain. Fulton's 1989 table has an avocado oil row, and it is a reading of the named material rather than a number borrowed from a neighbouring row. He printed it as "Avocado oil 3(2)." Three things travel with that number and are almost always dropped. First, Fulton's grade key: "a minimal grade of 0 to 1 is not considered significant. Grade 2 to 3 is borderline. However, a grade of 4 to 5 is uniformally reproduceable and considered positive." A 3 is the top of borderline — one step below a positive result, and not a positive result. Second, the figure in parentheses. Fulton defines the notation once, for the whole table: parentheses indicate results using refined oil. So the refined sample scored 2 — one grade lower, and still inside the borderline band. What the paper does not do is specify what the first sample was. It is not described as crude, virgin, cold-pressed or unrefined; it is simply the oil he tested, and the refined figure is the one he labelled. Third, the row sits in his Section VI, "Oils," and the asterisk in that heading is defined exactly once in the paper: "Results depend on source of raw material." It governs every oil and butter row he printed. Avocado is not merely covered by that caveat in the abstract — Fulton names it. In the body of the paper he writes that "among the natural oils such as sesame oil, avocado oil, and mink oil, the results are improved when a more refined oil is used." So inside a single paper avocado oil's number is already a range, 3 down to 2, and both ends of it are borderline. The other primary assay disagrees outright. Morris and Kwan, publishing in the Journal of the Society of Cosmetic Chemists in 1983 — six years before Fulton, at a different laboratory — ran their own rabbit-ear assay on 32 materials, on the same 0-to-5 scale, and scored Avocado Oil 0: the bottom of the "not significant" band. They were not deferential to Fulton; the same paper openly contradicts him on sodium lauryl sulfate and says so in print, so the 0 is an independent measurement rather than a copy of his, and the gap between it and his 3 is a real gap rather than a rounding difference. A shared 0-to-5 scale is not a shared protocol, though, and the two readings are not directly comparable. Fulton applied the ingredient at roughly 10% in propylene glycol to the inner ear of three rabbits; Morris and Kwan applied it undiluted in the ear canal over 14 applications and graded microscopically. Both papers print the words "avocado oil," and neither gives a species, cultivar, supplier, extraction method or refining state, so whether the two laboratories tested materially comparable oils cannot be established from the papers themselves. That is what the Caution rests on: one borderline legacy reading, dropping a grade for a refined sample, set against a zero from a second laboratory and against the human evidence discussed below. It is a precautionary classification taken from the higher of two conflicting animal results, not evidence that avocado oil is likely to clog pores. A reader who gave more weight to the independent, undiluted Morris and Kwan reading could reasonably file this oil as Clear, and we could not say they were wrong.

What the evidence doesn’t tell you

We identified no human comedogenicity study of avocado oil as an isolated ingredient whose methods and avocado-specific results we could read in full. The rating is built on two rabbit-ear assays from the 1980s — a model that scores follicular keratosis in a rabbit's ear, not acne on a person — that do not agree with each other. Fulton called his own 220-row survey "not at all definitive but simply designed to stimulate research," and warned that the rabbit ear is sensitive enough that "not everything that irritates this model will also irritate human skin." There are two pieces of human evidence, and neither settles the question. The first we could not read. Catambay, Villanueva and Verallo-Rowell presented a randomised, double-blind study using a modified Draelos human comedogenicity protocol as a poster at the American Contact Dermatitis Society's 2016 meeting, and it is reported in secondary descriptions to have rated avocado oil non-comedogenic on human skin while the same study's rabbit-ear arm rated it comedogenic. We could not retrieve the poster or a full abstract, so the avocado-specific result, the grade and concentration of oil used, the participant count and the funding and disclosure statements are all unverified here. The senior author is associated with a company that markets oil-based products on a non-comedogenic positioning; we could not establish who funded the work, and an affiliation is not by itself proof of sponsorship. Human skin evidence is generally more clinically relevant than a rabbit ear, but how much weight a particular human study carries depends on its directness, its quality, whether the underlying results can be inspected, and whether the ingredient was tested on its own. On those tests this one cannot yet be weighed at all. The second piece of human evidence is real and inspectable, and it is about formulas rather than about avocado oil. Waranuch and colleagues (2021) ran a double-blind randomised comedogenicity study in 15 healthy Asian male subjects with prominent follicular orifices, applying four finished dermatological products to the upper back over four weeks and counting microcomedones by epidermal biopsy under a stereomicroscope, against a threshold of a 50% increase. All four products stayed below that threshold; the positive control, octyl palmitate, exceeded it in every subject. Avocado oil was one ingredient among many in those formulas. A result about a product is not a result about an ingredient, and there is no way to read a score for avocado oil out of it. There is also human data on avocado oil and skin that measured something other than pores. The same 2022 Frontiers review concludes that in inflammation-affected skin, oils high in oleic acid and low in linoleic acid "may cause additional structural damage," while non-affected skin "may not react negatively to oils high in oleic acid." That is a barrier-function reading, not a comedone count: a weakened barrier is not a blocked follicle. The oleic-acid argument needs the same discipline. Avocado oil is roughly half to two-thirds oleic acid, and oleic acid is genuinely implicated in follicular trouble: Motoyoshi (1983) showed that free oleic acid — and, more strongly, its oxidation peroxides — induced sizeable comedones in rabbit-ear skin. But that experiment tested free oleic acid, not avocado oil. In avocado oil the oleic acid is bound into triglycerides, and "the fatty acid is comedogenic in isolation" is not the same claim as "the oil is comedogenic." A property of a component is not automatically a property of the material that carries it — the distinction this site exists to keep. Finally, "avocado oil" is not one fixed material. Cultivar, ripeness, and above all whether the oil is virgin or refined move both its fatty-acid profile and, in Fulton's own table, its score — 3 for the sample he tested, 2 for a refined one. A rating given across "avocado oil" is a rating given across a range, and the label rarely tells you which grade is in the bottle.

Where you’ll see it

On an ingredient list it is Persea Gratissima (Avocado) Oil. It is used as an emollient and skin-conditioning agent, so it turns up in the product categories those functions imply: facial and body oils, moisturisers and creams, cleansing and massage oils, balms and soaps. It is also a kitchen oil people apply to skin neat. Where it sits on an ingredient list tells you little on its own — grade and freshness, virgin versus refined and fresh versus oxidised, matter more to this page's question than label position does, and neither is printed on the label.

Sources

Each source says what it actually is. Several widely-cited “sources” for comedogenic ratings are republishing the same 1989 assay, and counting them as independent agreement is how a thin evidence base gets made to look thick.

1. Fulton JE. "Comedogenicity and irritancy of commonly used ingredients in skin care products." J Soc Cosmet Chem 40:321–333 (1989) The rabbit-ear assay nearly every online comedogenic score descends from, read here off a render of the journal scan rather than from a list quoting it. Method: ingredient at roughly 10% in propylene glycol, applied to the inner ear of three New Zealand albino rabbits, follicles scored for keratosis on a 0-to-5 scale with the key 0-1 "not significant," 2-3 "borderline," 4-5 "positive." It contains a real avocado oil row, printed "Avocado oil 3(2)," inside Section VI "Oils," whose heading asterisk reads "Results depend on source of raw material." The parenthetical is defined once for the whole table as the result using refined oil; the paper does not state the grade of the first sample. Fulton also names avocado oil in his body text as one of three oils whose results improve with refinement. A survey, on his own description "not at all definitive," reporting a borderline reading — not a single settled number.
2. 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 primary rabbit-ear assay of 32 materials on the same 0-to-5 scale as Fulton's and predating it by six years — not a republisher of him, and one that openly contradicts him in print on sodium lauryl sulfate. Its protocol is not Fulton's: material applied undiluted in the rabbit ear canal over 14 applications, graded microscopically only. It scored Avocado Oil 0. Cited as the second of the two primary rabbit-ear readings avocado oil has, and the reason its number is a disagreement rather than a finding. The score is taken from our verbatim transcription of the paper's oils table, which is printed rotated and is lost to OCR.
3. Catambay N, Villanueva M, Verallo-Rowell VM. "Comedogenicity of virgin coconut oil and other cosmetic oils using a modified Draelos protocol: a randomized double-blind controlled trial." Poster, American Contact Dermatitis Society Annual Meeting (2016) A conference poster — not a full peer-reviewed paper — reporting a randomised, double-blind study that is described as running a modified Draelos human comedogenicity protocol alongside a rabbit-ear arm, and as rating avocado oil non-comedogenic in humans while the rabbit-ear arm rated it comedogenic. We could not retrieve the poster or a full abstract, so the avocado-specific result is read from secondary descriptions and its methods, oil grade, concentration, participant count, funding and disclosures are unverified. The senior author is associated with a company that markets oil-based products on a non-comedogenic positioning; we could not establish the study's funding. Cited because it exists and points away from the Caution, not as a result we can weigh against the two inspectable assays.
4. Waranuch N, Wisutthathum S, Tuanthai S, Kittikun P, Grandmottet F, Tay F, Viyoch J. "Safety assessment on comedogenicity of dermatological products containing d-alpha tocopheryl acetate in Asian subjects: a double-blind randomized controlled trial." Contemp Clin Trials Commun 23:100834 (2021) A double-blind randomised human study of FINISHED PRODUCTS, not of an ingredient: four dermatological formulations applied to the upper back of 15 healthy Asian male subjects with prominent follicular orifices for four weeks, with microcomedones counted by epidermal biopsy under a stereomicroscope against a threshold of a 50% increase. All four products fell below the threshold; the positive control, octyl palmitate, exceeded it in every subject. Avocado oil was one of many ingredients in the tested formulas. Cited as the only inspectable human follicular evidence we found that involves avocado oil at all, and cited with the limit that governs it — it cannot isolate the oil, so it supports no score for it.
5. Poljšak N, Kočevar Glavač N. "Vegetable Butters and Oils as Therapeutically and Cosmetically Active Ingredients for Dermal Use: A Review of Clinical Studies." Front Pharmacol 13:868461 (2022) A peer-reviewed review of clinical studies of vegetable oils and butters, read here directly. Cited for two things only: avocado oil's published fatty-acid composition (oleic roughly 47-60%, palmitic 16-24%, palmitoleic about 7%, linoleic 13-14%, reported as ranges that vary with source), and its conclusion that in inflammation-affected skin, oils high in oleic and low in linoleic acid "may cause additional structural damage," while non-affected skin "may not react negatively to oils high in oleic acid." It is a barrier-function review, not a comedogenicity source, and it counts no comedones; used only to characterise the oil and to mark the line between "hard on the barrier" and "clogs pores."
6. Motoyoshi K. "Enhanced comedo formation in rabbit ear skin by squalene and oleic acid peroxides." Br J Dermatol 109(2):191–198 (1983) A rabbit-ear study of free oleic acid, squalene and others, together with their oxidation peroxides. Free oleic acid, and more strongly its peroxides, induced sizeable comedones, with comedone size tracking peroxide level. Cited for the oleic-acid mechanism behind avocado oil's reputation — and cited with its material-identity caveat stated: it tested free oleic acid, not avocado oil, in which the oleic acid is bound into triglycerides. Evidence about the component, not about the material.

Others in the same family

Oils behave similarly enough that the evidence for one is often wrongly read across to the others. These are its structural relatives, not a guess. What the evidence says about oils as a family ›

• Argan Oil Clear Low
• Castor Oil Clear Moderate
• Coconut Oil Clogging Moderate
• Corn Oil Caution Low
• Cottonseed Oil Caution Low
• Evening Primrose Oil Caution Low
• Flaxseed Oil Caution Low
• Grapeseed Oil Caution Low

Last reviewed 2026-08-01 · How we decide

‹ Check a full ingredient list