Preservatives: the most feared line on the label

Start with the physical situation. A sheet mask is a piece of fabric soaked in warm, nutrient-rich water, sealed in a sachet, and then stored at room temperature for eighteen months to two years. A clay mask is a wet paste in an open-topped jar that you put your fingers into repeatedly for six months. Both are excellent growth media.

Preservation is not an additive a brand includes to cut corners. It is the thing standing between you and applying a bacterial culture to your face for fifteen minutes, which is, in the small number of documented cases, exactly what has happened when preservation failed. Of everything on an ingredient list, the preservative is the entry doing the most unglamorous, most necessary work.

Which makes the state of public feeling about it strange: on our reading of these 215 labels, the preservative line is the one most likely to make a shopper put a mask back, and the one least likely to deserve it.

The fact that reframes this whole subject: in the EU, preservatives are governed by a positive list. Annex V of the Cosmetics Regulation names every substance that may be used as a preservative, with its maximum concentration and any conditions attached. If a substance is not on Annex V, it may not be used as a preservative in a cosmetic sold in the EU. Full stop.

Most ingredient regulation works the other way round: everything is permitted until it is restricted. Preservatives are the opposite, and it means the question “is this preservative safe?” has already been answered by a committee, in public, with a concentration attached, before the product reached the shelf.

masks name an Annex V preservative

103masks name an Annex V preservativeof 215

different Annex V preservatives used

11different Annex V preservatives usedacross the whole corpus

EU cap on phenoxyethanol

1%EU cap on phenoxyethanolthe most common one here, in 71 masks

You can read the whole annex yourself. It is a table, it is free, and it is the single most useful document for anyone who worries about this category: the consolidated Regulation 1223/2009.

The distribution is narrow. Two substances do most of the work across 215 masks, and the whole licensed set that appears here runs to 11 entries:

Phenoxyethanolis the workhorse: it holds across a wide pH range, is stable in the water-based essences sheet masks are built on, and has one of the lowest allergy rates of any preservative in routine use. It is capped at 1%. When France proposed cutting that to 0.4% for young children, the EU’s scientific committee reviewed the data and kept it at 1% for everyone.

112 of our 215 masks name no Annex V preservative at all. That sounds like a finding. It is mostly a labelling artefact, and unpacking it is the most useful thing in this piece.

• 93 use multifunctionals.Ingredients with real antimicrobial effect whose declared primary function is something else, so they are not “preservatives” in the legal sense, and a brand may truthfully say preservative-free while relying on them.
• 15 are preserved with alcohol. Ethanol at sufficient concentration is self-preserving, which is the classic natural-cosmetics route. Most of these are German natural-cosmetics brands, and it is why several of the most botanical formulas here carry the most alcohol.
• 4 are something else again: anhydrous powders and clays with no water for anything to grow in, or formulas where our matching has simply not resolved the relevant entry yet.

The alcohol-preserved group is worth looking at directly, because it is the clearest example of a trade-off being made rather than avoided: Balea Gesichtsmaske Glow Peel-Off, lavera basis sensitiv Anti-Falten Gesichtsmaske, Luvos-Heilerde Anti-Aging-Maske and others carry no licensed preservative and a volatile alcohol high in the list. If you avoid preservatives because your skin is reactive, that is not obviously the trade you wanted.

No ingredient family has had a worse two decades of public relations. Here is the state of the evidence, stated as plainly as we can manage.

Methylparaben appears in 6of these masks. It is capped at 0.4% as a single ester and 0.8% for parabens in total. It has been in use since the 1920s, which makes it one of the most thoroughly studied preservatives in existence, and it has a very low rate of contact allergy. The in-vitro oestrogen-receptor binding that started the scare is real and is orders of magnitude weaker than the body’s own oestradiol; skin and liver esterases break it down quickly. We grade it well tolerated.

Propylparaben appears in 1. Its longer side chain makes it more oestrogenically active in laboratory assays, and the EU treats it accordingly: the limit was cut from 0.4% to 0.14% in 2014, and it is not permitted in leave-on products for the nappy area of children under three. It remains on the EU’s priority list for potential endocrine-disrupting properties, so the limit could move again. We grade it worth knowing, not because 0.14% is dangerous, but because the question is open rather than closed.

What the differentiated treatment tells you

The interesting thing is not the verdict on either substance. It is that “parabens” is not a category with a single answer: two esters differing by one carbon chain get a 0.4% limit and a 0.14% limit, and two others, Isobutylparaben and Isopropylparaben, are banned outright. A label reading “paraben-free” collapses all of that into a marketing claim, and tells you nothing about what was used instead.

We deliberately document the preservatives with the worst reputations so that we can check for them. Across all 215 masks, the combined number of appearances is 0.

Methylisothiazolinone, the preservative behind the wave of contact allergy that made headlines across Europe in the early 2010s, is restricted out of leave-on cosmetics entirely and appears on none of these labels. The formaldehyde releasers appear on none. The banned parabens appear on none. This is a regulatory system that already did the work, and it is the reason a European mask aisle is a fundamentally different risk environment from an unregulated one.

One caveat we will not hide: absence from our corpus is not proof of absence from the market. We list 215 masks, not every mask in Europe, and our matching resolves about three-quarters of a typical printed list. Everything above is a statement about these labels.