Panthenol
Panthenol is rated Clear, and the number is attached to the right material for once — but the evidence is Fulton's rabbit table plus one sentence that the Cosmetic Ingredient Review carries forward from its own earlier report, summarising a single rabbit test at 0.5% we could not obtain. That 0.5% does not cover the reported use range, which reached 5.3% in one product category.
Humectant/soothing agent; non-comedogenic.
What it is
Panthenol is the alcohol analogue of pantothenic acid — vitamin B5 — and can be oxidised biologically to pantothenic acid. It is small, water-soluble and hygroscopic, which is why formulators use it as a humectant and skin-conditioning agent rather than as an emollient or an oil. The Cosmetic Ingredient Review's 2022 safety assessment lists its cosmetic functions as hair conditioning agent, skin-conditioning agent-humectant, and solvent; it is the only ingredient in that report's group with the skin-conditioning and solvent functions attached. Two forms are sold. D-panthenol (CAS 81-13-0), also called dexpanthenol, is the biologically active one; DL-panthenol (CAS 16485-10-2) is the racemic mixture, of which only the D half has vitamin activity. On a label you may see Panthenol, D-Panthenol, DL-Panthenol, Dexpanthenol or Provitamin B5, and which one you are looking at matters more to this page than it usually would — see the limits.
Why we rate it Clear
We rate panthenol Clear at Low confidence. The rating is not in doubt; the evidence behind it is thinner than the unanimity of every chart on the internet would suggest. The legacy number, and this time it is the right material. James Fulton's 1989 survey in the Journal of the Society of Cosmetic Chemists is the source of nearly every comedogenic score in circulation, and we read its Table I off a render of the journal scan rather than off a list quoting it. It prints a row for "Panthenol" at a comedogenicity grade of 0 on his 0-to-5 scale. The row sits in his vitamins-and-herbs section, between "Vitamin A palmitate" at 1-3* and "Methyl paraben" at 0. It carries no asterisk, and the section asterisk that governs every oil and butter row in the paper does not reach it. It is panthenol itself: the panthenyl derivatives — panthenyl triacetate, panthenyl ethyl ether — have no rows in the table at all. On Fulton's own published key, a grade of 0 to 1 "is not considered significant," so a 0 is the floor of the band he treated as noise rather than as a finding. A 0 from this assay deserves exactly the scepticism we give a 4 from it. The method is ingredient at roughly 10% in propylene glycol, 1 ml on the inner ear of three New Zealand albino rabbits, five days a week for two weeks, follicular keratosis measured under a micrometer. Fulton called his own survey "not at all definitive but simply designed to stimulate research." The other legacy rabbit surveys do not print it at all. Fulton's earlier 1984 survey with Pay and Fulton has no panthenol row, Morris and Kwan's 1983 table has none, and the 2007 Nguyen, Dang and Maibach rabbit study did not test it. There is one further line on the subject, and it is not a Fulton republisher — but it is not a second experiment either. The Cosmetic Ingredient Review's assessment of panthenol and its derivatives — final report, Panel meeting December 2017, released March 2018, published in the International Journal of Toxicology in 2022 — contains exactly one sentence on the question, tucked into its dermal irritation section: "Panthenol (0.5%) was non-comedogenic in rabbit skin." That sentence is footnoted to CIR's own earlier 1987 assessment, which in turn points at a contract report we could not obtain. Two documents carry the sentence between them; the experiment underneath is one experiment, summarised twice, and we have read neither its design nor its data. It is also, read carefully, very little. It is rabbits again, so it is not independent of the method, only of the paper. It reports a word, not a number, and no scale. The concentration is 0.5%, and CIR's own tables in the same report record a maximum reported use level of 5.3% in one product category, so the summarised test does not cover the reported use range. No design, no n, no endpoint definition is given. And the Panel does not carry the sentence into its summary, its discussion or its conclusion — it appears once and is never weighed. We tried to trace it and stopped short. The 1987 assessment (Elder RL, ed., Journal of the American College of Toxicology 6(1):139-162) is behind a paywall we did not get past; we read its abstract and its reference list. The abstract does not mention comedogenicity. The reference list contains exactly one comedogenicity entry: "IVY RESEARCH Laboratories, INC. (1983). Comedogenicity evaluation in the external ear canals of rabbits. CTFA Code No. 3-8-33" — a document identified by a trade-association code number rather than a journal, volume and page. We did not read it, and we are not asserting it is the source of the 2022 sentence; we are saying it is the only comedogenicity study in that bibliography and that the chain runs out there for us. So: one rabbit grade of 0, one adjective repeated across two documents, and a paper trail that ends at a file we cannot open. Clear is the right rating. Low is the right confidence.
What the evidence doesn’t tell you
We identified no human comedogenicity study of panthenol — no cyanoacrylate follicular biopsy, no occluded back application, no counted comedones on a person. Both readings are rabbit-ear readings. The rabbit ear is the model this site discounts when it condemns an oil, and consistency requires discounting it here: a 0 from a weak, oversensitive model is weak evidence of a 0. Two rabbit results are not two independent lines of evidence; they are the same method twice. Dose is not settled. The CIR sentence reports 0.5%. CIR's own use tables in the same report record panthenol at up to 5.3% in leave-on body and hand products. A single result at 0.5% does not contain a dose-response curve, and it is not a result at 5.3%. Neither document tells you which isomer was tested. Fulton prints "Panthenol" and nothing more. CIR states the problem against itself: "the panthenyl cosmetic ingredients are defined somewhat vaguely, without indication of stereochemistry. Unfortunately, much of the available literature is just as vague." Whether the material graded 0 was D-panthenol, DL-panthenol or an unspecified mix is not recoverable from either paper. A safety conclusion is not a comedogenicity finding. CIR concluded that panthenol is "safe in cosmetics in the present practices of use and concentration described in this safety assessment." That is a statement about safety, and this page does not borrow it as a statement about pores. One observation points the other way and belongs on the page rather than off it. CIR's report describes a 10-week randomised, double-blind trial in 207 women of a lotion containing 0.5% panthenol together with 4% niacinamide, 0.5% tocopheryl acetate, sunscreen, glycerol and other unspecified ingredients. Six treated subjects and two controls reported a mild transient burning sensation, and one treated subject reported dry skin and worsening of acne. That is a single self-reported adverse event in a multi-ingredient finished lotion, attributed to nothing by the Panel, and it is not evidence that panthenol clogs pores. It is the only line in either document that runs against the rating, which is exactly why it is printed here. Systemic vitamin B5 is a different question and will turn up first in a search. Saki and colleagues published a single-blind randomised trial in Dermatologic Therapy in 2025: 59 patients with mild to moderate acne, all on topical adapalene, half also given an intramuscular B5 preparation — the paper's title names pantothenic acid, its methods dexpanthenol — assessed at two months by the Global Acne Grading System, Investigator's Global Assessment and total lesion count. The injections added nothing over adapalene alone. That is a different route, a different question — treating acne someone already has — and a different endpoint from whether a topical humectant plugs a follicle. It neither supports nor undermines this rating. Finally, a rating belongs to an ingredient, not to a product. A formula that pairs panthenol with heavy occlusives can still clog pores; the 0 here belongs to the humectant.
Where you’ll see it
The INCI declaration is Panthenol. D-Panthenol and DL-Panthenol are stereochemical descriptors of that same ingredient; Dexpanthenol is the drug name, the INN for the D form, and belongs to the pharmaceutical register rather than the INCI one; Provitamin B5 is marketing language and not a declaration at all. All five turn up on packaging, and they do not tell you the same thing. CIR's 2022 report records 5,766 formulations containing panthenol in the FDA's Voluntary Cosmetic Registration Program as of 2017 — 3,543 leave-on and 2,178 rinse-off — with maximum reported concentrations of 5.3% in body and hand products, 5% in skin cleansing products and hair conditioners, 5% in baby shampoos, 3% in eye lotions and 2.5% in baby lotions, oils and creams. Those are a 2016 industry concentration survey and a 2017 registration count, not a current market survey, and a maximum reported use level is a formulation figure rather than a demonstration of tolerance across that range.
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.
- Fulton JE. "Comedogenicity and irritancy of commonly used ingredients in skin care products." J Soc Cosmet Chem 40:321-333 (1989) The rabbit-ear survey nearly every comedogenic number in circulation descends from, read off a render of the journal scan rather than off a list quoting it. Method: ingredient at roughly 10% in propylene glycol, 1 ml to the inner ear of three New Zealand albino rabbits, five days a week for two weeks, follicular keratosis graded under a micrometer on a 0-to-5 scale. Its Table I prints a row for "Panthenol" at 0, in the vitamins-and-herbs section, with no asterisk. Fulton's own key: 0 to 1 "not considered significant." Cited as the legacy score, for the correct material, and as the same contested assay this site discounts elsewhere — a reassuring 0 from it is as weak as a condemning 4.
- Scott LN, Fiume M, Bergfeld WF, Belsito DV, Hill RA, Klaassen CD, Liebler DC, Marks JG Jr, Shank RC, Slaga TJ, Snyder PW, Heldreth B. "Safety Assessment of Panthenol, Pantothenic Acid, and Derivatives as Used in Cosmetics." Int J Toxicol 41(3 suppl):77-128 (2022) A Cosmetic Ingredient Review expert-panel SAFETY assessment covering seven ingredients. Document state matters and this one is clean: status "Final Report," Panel meeting 4-5 December 2017, released 9 March 2018, published in the journal in 2022 — not the 2017 draft, which carries the "Distributed for comment only — do not cite or quote" footer. Cited for four things. One: the single comedogenicity sentence in the document, "Panthenol (0.5%) was non-comedogenic in rabbit skin," which sits in the dermal irritation section, is footnoted to CIR's 1987 assessment, gives no design or n, and is never repeated in the summary, discussion or conclusion. Two: the use data (5,766 formulations, 3,543 leave-on, maximum 5.3%). Three: the report's own statement that the panthenyl ingredients are "defined somewhat vaguely, without indication of stereochemistry" and that "much of the available literature is just as vague." Four: the single acne adverse event in a multi-ingredient lotion trial. A safety review is not a comedogenicity study, and we do not use its conclusion as one.
- Elder RL (ed). "Final Report on the Safety Assessment of Panthenol and Pantothenic Acid." Journal of the American College of Toxicology 6(1):139-162 (1987) The earlier Cosmetic Ingredient Review assessment that the 2022 report footnotes for its comedogenicity sentence. We read the abstract and the reference list only; the body is behind a paywall we did not get past. Cited for exactly two facts and no more. The abstract does not mention comedogenicity at all. The reference list contains one comedogenicity entry — "IVY RESEARCH Laboratories, INC. (1983). Comedogenicity evaluation in the external ear canals of rabbits. CTFA Code No. 3-8-33" — identified by a trade-association code number rather than a journal, volume and page, and not a document we or a reader can obtain. We did not follow the chain to the end and do not claim to have: we are not asserting that this 1983 report is the source of the 2022 sentence.
- Saki N, et al. "Efficacy of Intramuscular Pantothenic Acid in the Treatment of Acne Vulgaris: A Single Blind Randomized Clinical Trial." Dermatologic Therapy, article 6610699 (2025); doi:10.1155/dth/6610699 A single-blind randomised clinical trial in 59 patients with mild to moderate acne vulgaris. Both arms received topical adapalene gel; one arm additionally received an intramuscular injection, named as pantothenic acid in the title and as dexpanthenol in the methods — we print the discrepancy rather than resolve it. Endpoints at two months were the Global Acne Grading System, the Investigator's Global Assessment, total lesion count by the Leeds technique, and the Cardiff Acne Disability Index. The injections conferred no additional benefit. Cited ONLY to mark the boundary of this page: it is systemic administration, in people who already have acne, measuring treatment response. It measures no comedones caused by anything applied to skin, and it is not evidence about the comedogenicity of topical panthenol in either direction.
- CosDNA — panthenol ingredient entry An ingredient-lookup site listing panthenol at an acne rating of 0. Cited as one of the places a reader will have met the number, and characterised as what it is: a republisher of the legacy scores, not an independent measurement.
Others in the same family
Humectants 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 humectants as a family ›
Last reviewed 2026-08-12 · How we decide