Is mineral oil in skincare toxic, or does it clog pores?
Short answer: no, not the mineral oil actually used in cosmetics. The fear traces to two things that don't apply to what's in your moisturizer — industrial-grade mineral oil, which is a genuine occupational carcinogen, and a 1970s rabbit-skin test that later turned out to over-predict human pore-clogging. Cosmetic-grade mineral oil (INCI: paraffinum liquidum) is a different, far more purified material, and it's one of the most studied occlusives in dermatology.
What's actually the difference between "cosmetic-grade" and "industrial-grade" mineral oil?
Both start as the same petroleum distillate, but they're refined to very different endpoints, and the refining step is the whole story. The International Agency for Research on Cancer (IARC), the World Health Organization's cancer-classification body, splits mineral oils into two separate categories with two separate verdicts: untreated and mildly treated mineral oils are Group 1 — "carcinogenic to humans" — based on strong, consistent human epidemiological evidence of skin cancers, including scrotal cancer, in workers with heavy, prolonged skin contact with crude or lightly processed oils in trades like metal machining, mule-spinning, and jute processing. Highly refined mineral oils, by contrast, are Group 3 — "not classifiable as to carcinogenicity to humans" — because modern refining (severe hydrotreatment or solvent extraction) strips out the polycyclic aromatic compounds responsible for that risk [1].
Cosmetic-grade mineral oil (paraffinum liquidum, the ingredient profiled here) is required to meet pharmacopeial purity standards, not just "cosmetic" ones. In the EU, mineral oils are only permitted in cosmetics at all if the full refining history is documented and the starting material tests negative for carcinogenic potential using a specific method (IP346), or if it's been purified further until it does [2]. An independent government lab in Germany used NMR spectroscopy on 27 real-world samples — mostly near-pure mineral-hydrocarbon raw materials and pharmacopeial-grade products such as Vaseline, lip balm, and medicinal white oil, the closest available proxy for the purified material used in cosmetics — and found MOSH (the benign fraction) making up the large majority of each sample, with the aromatic fraction that carries carcinogenic risk (MOAH) mostly well under 1% (one raw-material sample read 1.10%) and no detectable PAHs in any sample tested [3]. That's the load-bearing distinction: "mineral oil" as an occupational hazard and "mineral oil" as a moisturizer ingredient are not the same material, and the safety data for one doesn't transfer to the other.
Does mineral oil in a moisturizer cause pores to clog?
The evidence for cosmetic-grade mineral oil specifically comedogenic is thin, and what looks like older confirmation is mostly an animal-model artifact, not a human finding.
Where the reputation came from
The origin point is Kligman and Mills' 1972 paper that coined the term "acne cosmetica." They tested dozens of cosmetic ingredients, including 23 petrolatum-type products, by applying them to the inside of rabbits' ears — a standard, but not perfect, comedogenicity screen — and found petrolatum (Vaseline) to be their most comedogenic petrolatum-type product in that model [4]. That's a rabbit-ear result, not a human one, and it's the finding that seeded decades of "petroleum-derived occlusives clog pores" advice.
Why the rabbit-ear finding didn't hold up in humans
Rabbit ear skin isn't a great stand-in for human facial skin — the follicles are anatomically different and the tissue reacts more readily, which produces comedones in the rabbit model that don't show up in people using the same substance. Later human testing on petrolatum specifically — applied to acne-prone skin over 8 weeks and checked by follicular biopsy — found no comedogenic effect at all, and by the late 1980s and into the mid-1990s, petrolatum's comedogenic rating had been walked back on this basis [5].
The most-cited direct evidence for cosmetic-grade mineral oil follows the same pattern. A 2005 paper reported that 100% cosmetic-grade mineral oil scored zero for comedogenic activity in the rabbit-ear assay, and that finished formulas containing up to 30% mineral oil tested non-comedogenic on human backs using a modified Kligman-Mills human assay [6]; a related 2006 review reached a similar conclusion and argued the rabbit-ear model as a whole over-predicts human pore-clogging [7]. Here's the disclosure that belongs next to that claim, not buried in a footnote: both papers share an author, Joseph DiNardo, a cosmetic chemist whose career includes stints at Revlon-Almay and Allergan and whose current consulting firm, Pharma Cosmetix Research, is a member of the Personal Care Products Council — the cosmetics industry's own trade association. That's a real commercial tie to the industry that sells mineral-oil-containing products, and both of the field's key rebuttal papers trace to the same person, which means "mineral oil is non-comedogenic" is not yet an independently-repeated finding — it's one consistent, industry-connected voice, not confirmed by a separate lab with no stake in the answer. The direction of the evidence — cosmetic-grade mineral oil testing low or zero on comedogenicity, and the rabbit model over-predicting generally — is plausible and consistent with what happened to petrolatum's rating, but it hasn't been re-run independently the way you'd want for a fully settled claim.
If the industrial-carcinogen data doesn't apply, why does the fear persist?
A few structural reasons, not because the underlying science is genuinely contested. First, consumer-facing hazard databases don't always separate refining grades cleanly — EWG's Skin Deep database, for instance, lists mineral oil and several of its petroleum-distillate variants at a "moderate hazard" tier, driven mostly by generic concerns like bioaccumulation and non-reproductive organ toxicity that its own methodology applies across grades, rather than a grade-specific finding for the purified cosmetic material [8]. A shopper scanning that score has no easy way to see that it's not describing the same substance IARC put in Group 3. Second, "mineral oil" and "petroleum" both carry an intuitive, non-technical association with industrial contamination that a chemical name like "dimethicone" doesn't trigger, even where the actual purity profiles are comparable. Third, the original comedogenic reputation is decades old and was never loudly corrected in consumer media the way it was in the dermatology literature — petrolatum's rabbit-ear reversal happened across the late 1980s and mid-1990s in trade journals most people never read.
Is mineral oil safe to use on skin every day?
For the highly refined, pharmacopeial-grade material used in cosmetics, current regulatory assessment says yes. Germany's Federal Institute for Risk Assessment (BfR), reviewing the dermal and lip-product exposure data most recently in 2018, concluded that no health risk is currently recognized from dermal or oral (lip-product) exposure to cosmetic-grade mineral oil at the purity levels required by EU cosmetics law, and noted there have been no reports of health effects from these ingredients despite decades of daily use in creams, lotions, and baby-care products [2]. One caveat worth stating plainly and labeling correctly: in animal studies, high doses of low-viscosity mineral oil produced inflammatory liver lesions (granulomas) specifically in female Fischer 344 rats — a rodent, high-dose, liver-specific finding — and BfR itself says the relevance of that result to humans is doubtful, given no corresponding signal has turned up in comparable human tissue with high mineral-oil exposure [2]. That's the honest state of the evidence: a mechanistically distinct industrial hazard that doesn't apply to the cosmetic material, a comedogenicity case that leans favorable but rests on industry-connected authorship, and no red flags from independent regulatory review of the purified ingredient itself. If you're weighing it against other occlusives in a slugging routine, or checking whether a "non-comedogenic" label on a product actually means anything, see what that claim is and isn't backed by.
