How to read this site

Three conditions, all of them hard: it has to be a face mask, it has to be buyable by a consumer in the EU right now, and its full ingredient list has to be published somewhere we can link to. A mask that fails the third condition does not appear, which is why some obvious names are missing.

The 215 masks here come from 89 brands across Korea, Japan and Europe, chosen to spread across formats (sheet, clay, wash-off, overnight, exfoliating) and price points rather than to flatter any brand. Japanese masks were the hardest: many are sold only in Japan, and several well-known ones were dropped because we could not confirm EU availability.

Every list was copied from a page that publishes it in full: the brand’s own site, a retailer’s ingredients tab, or an INCI database. Nothing here is retyped from a photograph of a box or assembled from a marketing page. We record where each list came from and the date we last checked it against that source.

We normalise capitalisation, because sources SHOUT inconsistently, and repair the damage retailer feeds do to a list: stray characters, an appended batch code, an obvious transposition like “Henxandiol”, a missing comma that welds two ingredients into one. Every such repair is recorded against the list it came from. We never reorder an entry or drop one.

We do not translate, and we will not use a translated list. Several EU retailers print the ingredients in German (“Wasser, Propandiol, Dikaliumglycyrrhizat”), which is a translation of a declaration, not the declaration. Where that was the only list available, the mask was left out, even when it was in stock and everything else about it checked out. Where a verbatim list existed elsewhere and matched the translated one entry for entry, we used the verbatim one. Where a source disagreed with another, we kept the verbatim list and recorded the conflict.

The main risk here is staleness. Brands reformulate without announcing it, and retailer ingredient tabs lag behind the pack. Treat what you see as accurate as of the date we last checked it, and read the actual box before you buy if you are avoiding something specific.

Ingredient lists are ordered by concentration, highest first, but only down to 1%. Below that, a brand may list in any order it likes. That single rule explains why a hero ingredient can sit second-to-last and still be on the front of the box.

We estimate where that line falls by looking for the first ingredient that is, in practice, always used at or below 1%: xanthan gum, carbomer, disodium EDTA, tocopherol, a preservative, a declared fragrance allergen. It is an estimate, not a disclosure, and we mark it as one. We found a defensible line on 201 of the 215 masks.

Every ingredient gets one of four levels. The rubric is deliberately calibrated rather than cautious: a level that says something is fine is as much a finding as one that says be careful. We are not in the business of scaring people off perfectly good formulas.

Well tolerated

No meaningful concern for most people; irritation is possible but rare.

Depends on you

Genuinely relevant for some people, or depends on dose and pH.

Worth knowing

EU-restricted, a common contact allergen, or subject to live debate.

Avoid

Banned in EU cosmetics or credibly linked to serious harm.

Concretely: parabens at EU-permitted concentrations are not toxic, and are graded accordingly. Silicones do not suffocate skin. Sodium hydroxide in a finished formula is a pH adjuster, not a caustic exposure. Where a genuine restriction, allergy risk or contamination question exists, we name it and link the regulation.

Benefits carry an evidence meter: four ticks for strong, one for traditional use. It is shown in ink rather than colour on purpose, so it can never be mistaken for a safety signal.

• Strong: repeated controlled human trials, or a regulatory body’s own assessment.
• Moderate: at least one decent human study, or consistent smaller ones.
• Limited: in-vitro, animal, or small uncontrolled human work only.
• Traditional: long use and plausible mechanism, little published testing.

299 sources are cited across the encyclopaedia. Every one links to the original: a journal, a PubMed record, an SCCS opinion or an EU regulation. We do not cite ingredient-scoring apps or brand pages as evidence.

The notes on a product page (“contains fragrance”, “3 declared fragrance allergens”, “drying alcohol, high in the list”) are computed from the ingredient list by matching it against the encyclopaedia. Nothing is written per product by hand, so a brand cannot earn a nicer summary by having nicer marketing.

Position matters and is used: alcohol at #3 is flagged differently from alcohol at #22, because the first tells you it is most of the formula and the second does not.

Matching is imperfect. On average we resolve 77% of a list to an encyclopaedia entry; the rest are shown as printed, with a dot instead of a rating. An unmatched ingredient is not a suspicious one; it just means we have not written it up yet.

There are no affiliate links on this site, no sponsored placements and no paid inclusion. Retailer links are plain URLs with no tracking parameters. Nothing about a brand’s commercial relationship with anyone affects its position in a list, because there is no relationship.

Prices are indicative, captured from the source recorded on each product page, and will drift. We show price per use as well as pack price, because a €25 jar that lasts twenty uses is cheaper than a €4 sheet mask.

Concentrations. Brands do not publish them, and we do not guess. An ingredient list tells you what is in a product and roughly in what order, not how much. A mask with salicylic acid could be at 0.1% or at the EU maximum.

Whether it will work for you. Nothing here is a review. We have not tested these masks on skin, and a formula that reads well can still sting. Patch-test anything new, especially with actives or essential oils.

Reformulations. See above: lists go stale, and we will not always catch it first.

Product photography comes from brand and retailer sources for identification; a couple carry a retailer watermark. They are the manufacturers’ images, not ours.

Found something wrong? That is the most useful thing you can send us. Ingredient corrections with a source link get fixed fastest.