Porecelain

Flawless skin starts with the fine print.

Phenoxyethanol

Phenoxyethanol is rated Clear, and the 0 out of 5 that follows it around the internet is not a measurement of it. Fulton's 1989 table, the source of most comedogenic numbers in circulation, has no phenoxyethanol row. What it has, sitting between methyl paraben and propylparaben, is phenoxyethyl paraben at 0 — a different molecule. We identified no comedogenicity assay of phenoxyethanol in any of the primary documents we hold. The modern safety literature on this ingredient is unusually large, and none of it asks the question.

Clear Low confidence No legacy score. It does not appear in the assay corpus we searched.

No assay we can find. Fulton's nearby 0 is phenoxyethyl PARABEN, a different material.

What it is

2-phenoxyethanol (CAS 122-99-6), a glycol ether: a benzene ring joined by an oxygen to a two-carbon chain ending in an alcohol. It is a broad-spectrum preservative, effective against Gram-negative and Gram-positive bacteria and yeasts, with only a weak inhibitory effect on resident skin flora, which is why it is normally paired with a second preservative rather than used alone. In the European Union it is authorised under Annex V, entry 29 of Cosmetics Regulation (EC) No. 1223/2009 at a maximum of 1%. It is the preservative most formulations commonly reach for when they are sold as paraben-free — a pattern we have observed on labels, not a market share we have counted. That last phrase is worth holding on to, because the argument on this page turns on a name. PHENOXYETHYL PARABEN is not phenoxyethanol. A paraben is an ester of 4-hydroxybenzoic acid, and phenoxyethyl paraben is the one whose alcohol component is phenoxyethanol — the phenoxyethyl group bonded to a hydroxybenzoate. It is a larger molecule, of a different chemical class, from the family whose other members are methylparaben and propylparaben. The two names differ by one word. The materials do not.

Why we rate it Clear

We rate it Clear at Low confidence, and the second half of that sentence is doing as much work as the first. Start with the table. We read all 220 rows of Fulton's 1989 Table I off a 400dpi render of the journal scan. There is no phenoxyethanol row in it. There is no phenoxyethanol row anywhere in the paper. What the table does contain, in a run of preservatives, is three consecutive entries: "Methyl paraben — 0", "Propylparaben — 0", "Phenoxyethyl paraben — 0". By Fulton's own published key a grade of 0 to 1 "is not considered significant," so those are three low scores for three parabens, and the third of them is not this ingredient. Now the claim we are NOT going to make. It would be easy, and satisfying, to write that the chart makers misread that row — that the 0 the internet gives phenoxyethanol was lifted off the phenoxyethyl paraben line. The names are one word apart and the numbers match, and our own reference file records the suspicion. But a matching number is not a demonstrated citation chain. We did not follow the trail from that row to any published chart, and until someone does, "they copied it" is a story that fits the evidence rather than a finding. What we can state is narrower and still worth the page: the primary table has no row for this ingredient, and it has a row for a near-namesake that is a different molecule. An earlier draft of this page went further and speculated that Fulton's "phenoxyethyl paraben" might be a garbled printing of phenoxyethanol. We have removed that, because we have no evidence of a printing error and we do not get to invent one to make a number usable. What the table prints is what the table prints. And even granting the most charitable reading anyone could ask for, it would still be a single unconfirmed 1989 rabbit-ear 0, sitting in a paraben cluster, from a survey its own author called "not at all definitive but simply designed to stimulate research." Either way the number is not a foundation. We think it is a paraben, because that is what the paper prints and because that is the company it keeps in the table, and we are telling you the alternative because the honest position is that one row cannot settle this. We checked the second primary assay too. Morris and Kwan, 1983, six years before Fulton, in a different laboratory and on the same 0–5 scale: we hold their rows, and phenoxyethanol is not among them. Our reference file records it in the untraceable set — the ingredients for which no primary document in our corpus assigns a score. What exists instead is a large modern safety literature, and it is the most interesting thing here, because of what it is not. The European Scientific Committee on Consumer Safety issued a final opinion in 2016 concluding phenoxyethanol is safe as a preservative at up to 1% for all consumers including children of all ages. The Cosmetic Ingredient Review reached a comparable conclusion in its 1990 final report and reaffirmed it in 2007. Dréno and colleagues published a review of the whole safety picture in the Journal of the European Academy of Dermatology and Venereology in 2019. This ingredient has been examined for dermal absorption, systemic toxicity, genotoxicity, carcinogenicity, reproductive and endocrine effects, eye irritation, skin irritation and contact allergy. None of it is about pores. We searched the full text of the Dréno review: the words comedogenic, comedone, acne, pore and follicle do not appear in it. Not in the abstract, not in the section on local effects, not in the tables, not in the reference list. The SCCS opinion and the CIR report are safety assessments, and a safety assessment is not a comedogenicity measurement. An ingredient can be studied exhaustively for one question and never once for the other, and phenoxyethanol is the cleanest example of that on this site. So read the rating precisely. Clear, here, means no comedogenic concern is on record — and none is. It does not mean the ingredient was tested and found inert. It was not tested. The rating is an inference from an absence, and the confidence is Low because that is what an inference from an absence is worth.

What the evidence doesn’t tell you

We identified no comedogenicity assay of phenoxyethanol in the assay corpus we searched (listed at /how-we-decide#corpus, and treated as OPEN — every time we have said 'there are only N documents', N has been wrong), and none in the searches we ran for this page. That is a bounded statement about what we looked at, not a claim about the world's literature, and it is the strongest form the evidence supports. The concentration gap is real and we are not going to reason across it. The legacy assays apply their test material at roughly 10% in propylene glycol. Phenoxyethanol is capped at 1% in the EU and is typically used at or below that. Had a rabbit-ear number existed it would have been generated at a concentration an order of magnitude above anything a product contains — but there is no number, and we have no dose-response curve for a material that does not appear in the corpus we searched, so this observation cuts no further than saying the assay and the product were never in the same place. What IS documented about this ingredient on skin is allergy, not plugging, and the distinction is the whole point. As reported in the Dréno review, patch-test surveillance in Germany, Switzerland and Austria found positive reactions to 1% phenoxyethanol in 0.24% of 6932 subjects, and 0.1% in an earlier cohort; the authors concluded it could be regarded as an extremely rare allergen, if an allergen at all. A handful of contact urticaria and contact dermatitis case reports exist. We have not read those primary papers ourselves and are reporting them as the review reports them. And a rare contact allergen is a rare contact allergen: it tells you nothing whatever about follicular keratosis. Allergy is not comedogenicity, irritation is not comedogenicity, and dermal-penetration toxicokinetics — of which there is a good deal for this molecule — is not comedogenicity either. The review we lean on hardest has a disclosure that belongs on the page. Dréno et al. 2019 is a narrative review, not a systematic one, published in a JEADV supplement. Its medical writing was funded by Cosmétique Active International, L'Oréal's active-cosmetics division; two of its authors sit on that company's scientific advisory board, and three L'Oréal R&D staff are credited with critically reviewing the manuscript. Its literature search ran PubMed only, over a window from November 2015 to May 2019, inheriting everything earlier from the SCCS opinion, and a large share of its underlying evidence is unpublished industry study reports. We cite it mainly for what it does not contain, which is a use its funding does not compromise — but a reader deserves to know who paid for it. Two things about the number you will have seen, and neither is a source. AI-written ingredient pages will hand you "comedogenic rating 0/5, irritancy 1/5" for phenoxyethanol with no author, no study and no scale named. We found several while researching this page and used none of them. And acne.org, which people reasonably treat as authoritative, does not rate phenoxyethanol 0: its list is 45 ingredients it judges LIKELY COMEDOGENIC, compiled from 11 studies, and phenoxyethanol is simply absent from it. Absence from a list of comedogenic ingredients is not a score of zero. It means the ingredient was not flagged, which is a much weaker statement, and turning the one into the other is exactly the move this site exists to refuse. One more gap, and it is ours. The entry in our own ratings file cites CosDNA as its only source. We tried to read CosDNA's page for this ingredient while writing this and could not retrieve it, so we are not going to tell you what CosDNA prints. That is a hole in our sourcing, and it is named in the rating challenge on this page rather than quietly left alone. Finally, the field-wide caveat, which applies with unusual force to an ingredient that has no number at all: a 2025 review in JAAD Reviews surveying comedogenicity testing from 1972 onward found that no standardised comedogenicity test exists and that the label "non-comedogenic" is unregulated. There is no procedure this ingredient could have been put through that would have settled the question cleanly, even if someone had tried.

Where you’ll see it

Very widely, and almost always at or below 1%, which is the ceiling the EU sets. It appears in leave-on and rinse-off products across essentially every category, and it is the usual preservative in formulations marketed as paraben-free. Those are categories, not a measured market share — we have not audited the market, and we are not going to tell you where it sits on a label, because label position is not a dose. One regulatory note a reader may meet: the French agency ANSM recommends against phenoxyethanol as a preservative in products intended for the nappy area of children under three, including wipes. That recommendation follows from a worst-case aggregate exposure calculation, and it is a systemic exposure judgement. It has nothing to do with pores.

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 primary source of most comedogenic numbers in circulation, read off a 400dpi render of the journal scan rather than via a chart quoting it. Method: test material at roughly 10% in propylene glycol, applied to the inner ear of New Zealand albino rabbits, three rabbits per assay, five days a week for two weeks, follicular keratosis measured under a micrometer. Fulton's own key: 0–1 "is not considered significant", 2–3 borderline, 4–5 "considered positive". He described the survey as "not at all definitive but simply designed to stimulate research". CITED HERE FOR AN ABSENCE: all 220 rows of Table I contain no phenoxyethanol entry, and neither does any other table in the paper. The near-namesake it does contain is "Phenoxyethyl paraben — 0", printed immediately after "Methyl paraben — 0" and "Propylparaben — 0" — a paraben, in a run of parabens, and not this ingredient.
  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 on the same 0–5 scale, six years BEFORE Fulton 1989 — not a republisher of him, and it contradicts him in print elsewhere. Cited here purely for what it does not contain: we hold its rows, and phenoxyethanol is not among them. It is the second of the two full row-lists we can read directly, which is why an absence in it is worth stating.
  3. Scientific Committee on Consumer Safety (SCCS). "Opinion on Phenoxyethanol in cosmetic products" — SCCS/1575/16, final version of the opinion, adopted 6 October 2016 A regulatory SAFETY assessment by the EU's expert committee, and the basis of the 1% ceiling in Annex V of Regulation (EC) No. 1223/2009. Document state named deliberately: this is the FINAL version of the opinion, not a draft or a preliminary version released for comment. It concluded phenoxyethanol is safe as a preservative at up to 1% for all consumers including children of all ages. It assesses systemic toxicity, absorption and margins of safety. IT IS NOT A COMEDOGENICITY STUDY and we do not use it as one — safety and comedogenicity are different questions, and this document only answers the first.
  4. Dréno B, Zuberbier T, Gelmetti C, Gontijo G, Marinovich M. "Safety review of phenoxyethanol when used as a preservative in cosmetics." J Eur Acad Dermatol Venereol 33(Suppl 7):15–24 (2019) A NARRATIVE review (not systematic) in a JEADV supplement, and the most complete single account of what is known about this ingredient's safety. Its search ran PubMed only, from November 2015 to May 2019, inheriting all earlier evidence from the SCCS opinion; much of the underlying data is unpublished industry study reports. FUNDING, DISCLOSED IN FULL: medical writing was funded by Cosmétique Active International, L'Oréal's active-cosmetics division; authors Dréno and Gontijo sit on that company's scientific advisory board; three L'Oréal R&D staff are acknowledged for critically reviewing the manuscript. CITED PRIMARILY FOR WHAT IT DOES NOT CONTAIN. We searched its full text: the words comedogenic, comedone, acne, pore and follicle do not occur anywhere in it. Its endpoints are dermal absorption, repeated-dose toxicity, genotoxicity, carcinogenicity, reproductive and endocrine effects, neurological effects, eye irritation, skin irritation and contact allergy. It is also the source of the patch-test figures quoted on this page (0.24% of 6932 subjects, 0.1% of an earlier cohort, from IVDK surveillance) — figures we are reporting as the review reports them, having not read those primary papers ourselves.
  5. Cosmetic Ingredient Review. "Final Report on the Safety Assessment of Phenoxyethanol." J Am Coll Toxicol 9(2):259–277 (1990); use and concentration re-reviewed and reaffirmed 2007 The US expert-panel SAFETY assessment. Concluded phenoxyethanol safe in the present practices of use and concentration, generally below 1%, and found it neither a primary nor a cumulative skin irritant, nor a sensitizer, nor phototoxic. The panel reviewed newer data in 2007 and did not reopen the assessment. An irritancy finding is not a comedogenicity finding, and this document reports no comedogenicity result. Cited as a safety document, and only as one.
  6. acne.org — "What Is Comedogenicity, and What Ingredients Are Comedogenic?" Not a scored index of every ingredient, which is how it is usually treated. It is an INCLUSION LIST: 45 ingredients acne.org judges likely to be comedogenic, compiled by its own account from 11 studies dated 1972 to 2009 (Kligman and Mills, Morris and Kwan, Fulton 1984 and 1989, Lanzet, Nguyen and Maibach, and others), with a two-tier rule — "avoid" if two rabbit assays or one human method flagged it, "consider avoiding" if one rabbit assay did. Phenoxyethanol does not appear on it. Cited to make one distinction precise: absence from a list of comedogenic ingredients is not a measured zero, and we are not going to present it as one.
  7. "Comedogenicity in cosmeceuticals — a review of clinical relevance, regulatory gaps, and future directions." JAAD Reviews (2025) A recent peer-reviewed review of the whole field, covering the literature from 1972 onward, independent of Fulton and of the charts that republish him. Finds that no standardised comedogenicity test exists and that the term "non-comedogenic" is unregulated. Cited for the field-wide caveat, which bears on this page with particular force: there was no settled procedure phenoxyethanol could have been put through even if someone had set out to test it.

Others in the same family

Preservatives 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 preservatives as a family ›

Last reviewed 2026-07-13 · How we decide

‹ Check a full ingredient list