DMDM Hydantoin and Formaldehyde Releasers: What They Are and Why They're On Your Shampoo Label

You lean over the side of the bath and pick up your own bottle. Bright label, professional-looking name, the brand your mum has always bought. Turn it around. Eight ingredients in. There it is.

DMDM hydantoin is a cosmetic preservativeA chemical added to a personal-care product to prevent the growth of bacteria, mould and yeast over the product's shelf life. Without preservatives, water-based cosmetics develop visible contamination within weeks. that has been on shampoo and body-wash labels in mainstream brands since at least the 1980s — its first CIRCosmetic Ingredient Review safety report dates to 1988 CIR 1988. It works by slowly hydrolysing in the bottle and on the skin to release CH₂O — formaldehyde — which does the actual antimicrobial work. IARCInternational Agency for Research on Cancer classifies formaldehyde itself as a Group 1 human carcinogen, on the basis of long-term occupational exposure in embalmers and plywood plants. The consumer story is different in kind — it's primarily about allergic contact dermatitis at consumer concentrations, not cancer at industrial ones — and we cover the cancer arc in our profile of formaldehyde itself. This article is about the indirect exposure route most consumers don't know is on their labels.

DMDM hydantoin (C₇H₁₂N₂O₄, CAS 6440-58-0) is the cosmetics-industry name for 1,3-bis(hydroxymethyl)-5,5-dimethylhydantoin — a preservative whose central job is to release small amounts of formaldehyde into a product over time. The molecule has two N-hydroxymethyl groups, each of which can cleave off via hydrolysis to give one free formaldehyde molecule plus the dimethylhydantoin core. Picture two doses of formaldehyde wrapped around a central scaffold like flags on a kite, each one falling off in turn as the bottle warms up in the shower. Two equivalents of formaldehyde per molecule, released gradually as the bottle sits on a shelf and as the product spreads on warm wet skin. That's what the 'releaser' designation actually means.

The original CIR safety assessment in 1988 concluded that DMDM hydantoin was safe for use as a cosmetic preservative at concentrations up to 1%, with the active ingredient existing in equilibrium with about 2% free formaldehyde in solution CIR 1988. The EU's current preservative cap under Regulation 1223/2009The EU Cosmetics Regulation, which sets a positive list of permitted preservatives in Annex V along with maximum concentrations and conditions of use. is 0.6% in finished products. The chemistry that the CIR panel signed off on in 1988 is the chemistry on the back of your shampoo bottle now.

What changed isn't the molecule. It's how much we know about how the immune system responds to repeated low-dose formaldehyde exposure on broken or barrier-disrupted skin — and how visible that science has become outside dermatology clinics.

Five names account for the vast majority of consumer exposure to formaldehyde-releasing preservatives: DMDM hydantoin, quaternium-15, imidazolidinyl urea (trade name Germall 115), diazolidinyl urea (Germall II) and bronopol — full name 2-bromo-2-nitropropane-1,3-diolA nitro-alcohol preservative that releases formaldehyde under alkaline conditions or at elevated temperatures. Also of separate concern: bronopol's decomposition products can react with secondary amines and amides commonly found in cosmetic formulations to produce nitrosamines.. Two minor releasers — sodium hydroxymethylglycinate and 5-bromo-5-nitro-1,3-dioxane (bronidox) — turn up less often. MethenamineA formaldehyde donor better known as a urinary-tract antiseptic in oral medicine. In cosmetic formulations it releases formaldehyde under acidic or heated conditions. is occasionally used in dermatological preparations.

The releasers don't release formaldehyde at the same rate. Lv and colleagues at Beijing Technology and Business UniversityA research group based in Beijing whose 2015 study used HPLC-DAD with 2,4-DNPH derivatisation to measure formaldehyde release from cosmetic preservatives in matched aqueous and finished-product matrices. measured release from each in 2015 using high-performance liquid chromatography after 2,4-DNPH derivatisation. The order, from highest release to lowest, ran: paraformaldehyde, then diazolidinyl urea, then DMDM hydantoin and quaternium-15 and imidazolidinyl urea (roughly tied), then methenamine, then bronopol, with a polymer-bound releaser called poly(p-toluenesulfonamide-co-formaldehyde) at the bottom Lv et al. 2015. Release was strongly matrix-dependent — water releases more than the formulated product, alkaline more than acidic, warmer more than cooler. A bottle sitting on a hot bathroom shelf releases differently to one in a cool cupboard.

Until 2010, the imidazolidinyl urea trade name Germall 115 was the most-used formaldehyde-releaser in both US and European cosmetics, ahead of DMDM hydantoin in second place de Groot & Veenstra 2010. Fifteen years later that ranking has flipped. Washington State's Department of Ecology pulled the Mintel Global New Products DatabaseA commercial database tracking new product launches and ingredient declarations across global consumer-goods categories. Used by regulators to estimate market prevalence of specific ingredients. in 2025 as part of its rule-making for the Toxic-Free Cosmetics Act and found that, of US cosmetic products containing any formaldehyde-releaser, DMDM hydantoin appeared in 44% of US cosmetic products containing any formaldehyde-releasing preservative — Washington Department of Ecology / Mintel 2025 of them — top of the prevalence ranking, ahead of diazolidinyl urea (32%) and imidazolidinyl urea (14%).

The strongest body of evidence on formaldehyde releasers at consumer concentrations is dermatological, not oncological. Allergic contact dermatitis is the documented endpoint: an immune-mediated reaction in which the skin becomes chronically inflamed on re-exposure to a chemical the immune system has previously marked as foreign. Once your immune system marks formaldehyde — and the releasers register as formaldehyde to the Langerhans cellThe skin's primary antigen-presenting immune cell, located in the epidermis. When a small reactive molecule like formaldehyde binds to a self-protein in the skin, the Langerhans cell processes it and presents the modified protein to T cells in the lymph nodes, which is the start of contact sensitisation. machinery in the skin — the marking is essentially permanent. Subsequent contact triggers the reaction. Cumulative low-dose exposure is the typical induction route, which makes daily personal-care products mechanistically the most plausible vector.

The cleanest population-level number comes from the North American Contact Dermatitis GroupA consortium of 12 academic dermatology centres across the US and Canada that has run standardised patch-testing on referred patients since 1994 and pools data biennially. Patch-tested patients are pre-selected as suspected contact dermatitis cases, so the percentages are higher than would be expected in the general population — but the trends and inter-allergen rankings within the cohort are reliable.. Atwater and colleagues pooled the group's records from 50,799 patch-tested patients across 1994 through 2016 and reported the per-preservative reaction rates: 7.8% of 50,799 patch-tested patients reacted to formaldehyde — NACDG pooled data 1994-2016, Atwater 2021 reacted to formaldehyde itself, and 7.7% of the same cohort reacted to quaternium-15 — the most common formaldehyde-releaser in the patch-test panel to quaternium-15 Atwater et al. 2021. Diazolidinyl urea ran at 2.8%, bronopol at 2.5%, imidazolidinyl urea at 2.4% and DMDM hydantoin at 2.0%. Methylisothiazolinone — a non-formaldehyde-releasing preservative we'll come back to — topped the list at 12.2%.

The qualifier matters. NACDG patients are referred to specialist dermatology clinics because they're already suspected of having contact dermatitis. The 7.8% rate is not a general-population rate. It is the rate within a population pre-selected for likely sensitisation. And it tells you that of every thirteen people the dermatologist tests, one is sensitised for life. European patch-test cohorts on less-selected populations report formaldehyde sensitisation rates closer to 1-3% de Groot 2010 Part 2 — lower than NACDG's referred cohort, but not negligible at population scale.

In 2015, the American Contact Dermatitis Society named formaldehyde its Allergen of the Year, an annual designation chosen to direct clinical attention to under-recognised everyday exposures Pontén & Bruze 2015. The designation matters less for the clinicians who already know than for the consumers who don't. Formaldehyde is the only preservative-class allergen named twice in the society's history, by way of methylisothiazolinoneA preservative widely used as a replacement for formaldehyde-releasers from the late 1990s onward, particularly in leave-on cosmetics. Triggered an epidemic of allergic contact dermatitis through the 2000s and 2010s., the chemical the industry pivoted to after parts of the cosmetic preservative market began moving away from formaldehyde-releasers in the 2000s. We'll come back to that — it's the regrettable-substitution piece of this story.

It comes up in every TikTok thread. The short version: IARC classifies formaldehyde itself as Group 1 — confirmed carcinogenic to humans — based on long-term occupational exposure in embalmers, plywood plant workers and chemical-industry employees, with sufficient evidence for nasopharyngeal cancer (2006) and leukemia (2012). Consumer-level exposure from a shampoo containing 0.6% DMDM hydantoin is orders of magnitude lower than what the embalming-industry cohorts breathed, and the cancer evidence base for releasers in personal care products specifically is much weaker than the dermatological one.

The strongest consumer-product cancer signal in the broader formaldehyde literature isn't releasers in shampoo — it's chemical hair-relaxer use, where the NIH Sister Study and the Black Women's Health Study both linked frequent use to elevated uterine cancer risk in cohort data. Hair relaxers are formulated, applied and heat-activated very differently to a daily shampoo, and they contain a stack of chemicals beyond formaldehyde and its releasers. We narrate the relaxer epidemiology in detail in our profile of formaldehyde — for releasers in personal-care products at consumer concentrations, the better-documented endpoint is the one your dermatologist sees: contact dermatitis, possibly for life.

The framing matters. If you find yourself reading 'shampoo and cancer' headlines, the evidence that exists is on long-term occupational inhalation in industrial settings, not on a 0.3% DMDM hydantoin rinse-off product. The reasonable consumer-side question isn't oncological at all. It's whether daily exposure to a slow-release formaldehyde preservative is the kind of cumulative low-dose contact a sensible person would choose to avoid when alternatives exist at the same price point. That's a different question, and the answer to it is mostly about your skin.

Parts of the industry did. From the late 1990s through the 2010s, formulators in Europe and increasingly the US pivoted away from formaldehyde-releasers — particularly in leave-on cosmetics — and substituted in MImethylisothiazolinone and the related MCImethylchloroisothiazolinone. The replacements were marketed as gentler. The new preservatives were used at much lower concentrations than the releasers, often at parts-per-million scale, which on the regulatory paperwork looked like a clear win.

What happened next is in the patch-test record. By 2013, the American Contact Dermatitis Society had named methylisothiazolinone its Allergen of the Year — five years ahead of European regulators getting around to restricting it in leave-on products Castanedo-Tardana & Zug 2013. The Atwater 2021 NACDG pooled data shows MI sensitisation hitting 12.2% of patch-tested patients across the cohort, higher than formaldehyde itself, and the Houle 2025 update shows the rate continuing to decline since the EU's 2017 leave-on restriction but still elevated Houle et al. 2025. The replacement preservative produced a larger contact-allergy epidemic than the chemical it replaced.

The pivot back to formaldehyde-releasers — not in Europe's leave-on cosmetics, but increasingly in US rinse-off shampoos and conditioners — is part of why DMDM hydantoin is now first by prevalence in US cosmetic releasers, well above the imidazolidinyl urea that led the 2010 ranking. The system that picks preservatives optimises for product stability and patent position. Skin sensitisation in the population takes longer to show up than mould growth in a bottle.

There is a counter-evidence position worth naming. An industry-funded quantitative risk assessmentA regulatory toxicology methodology that calculates a margin of safety by comparing estimated consumer dermal exposure to a no-effect dose for sensitisation derived from clinical or animal data. Widely used in the cosmetics and personal-care industry to defend ingredient use at typical concentrations. of DMDM hydantoin in shampoo published in 2023 concluded that, at allowable use concentrations and rinse-off contact times, the margin of safety was greater than 1 and that shampoos containing DMDM hydantoin were not expected to increase the risk of skin sensitisation induction Stewart et al. 2023. The Stewart 2023 paper is an industry-funded QRA — it should be read as part of the regulatory case for permitting DMDM at the current cap, not as independent epidemiology of consumer outcomes. The dermatology literature, working from observed patch-test rates rather than modelled exposure scenarios, reaches a more cautious view.

The regulatory picture for formaldehyde releasers is fragmented, fast-moving in some jurisdictions and stuck in others. The EU has tightened the labelling threshold by a factor of 50. Three US states have moved on cosmetic preservatives in the last five years, and a fourth has moved much further than the others. The federal US regulator has missed every deadline it has set itself.

The EU's biggest move on releasers wasn't a ban. It was a labelling threshold. Under EU Reg 2022/1181 2024 the concentration of free formaldehyde at which a product must carry the words 'releases formaldehyde' on the label was cut from 0.05% (500 parts per million) to 0.001% (10 parts per million) — a fifty-fold tightening. The new threshold has applied to products placed on the market since 31 July 2024, with a sell-through window for products already in distribution running until 31 July 2026. The change was driven by an SCCS opinion (1632/21) which concluded that the previous threshold did not protect formaldehyde-sensitised consumers.

California's Toxic-Free Cosmetics Act, US AB 2762 2020, took effect on 2025 and was the first US state ban of cosmetic ingredients. It prohibited 24 substances including formaldehyde, paraformaldehyde, methylene glycol and quaternium-15. It did not ban DMDM hydantoin, imidazolidinyl urea, diazolidinyl urea or bronopol. The follow-up bill AB 496, signed in 2023 and effective 2027, added 26 further substances — none of them additional formaldehyde releasers. California's statute is the textbook case of regulating chemicals by individual name rather than by mechanism. Maryland's HB 643 followed the California model the next year. The two states' lists are nearly identical.

Washington went further. US HB 1047 2023 authorised the state's Department of Ecology to regulate cosmetic preservatives as a class. The Department's resulting rule — Chapter 173-339 of the Washington Administrative Code, adopted in late 2025 — prohibits 28 formaldehyde-releasing chemicals when intentionally added to a cosmetic product, with DMDM hydantoin at the top of the list by prevalence. The substantive prohibition takes effect on 1 January 2027, with retailers permitted to exhaust existing stock through 1 January 2028. Washington became the first US state to regulate formaldehyde-releasers by their shared mechanism rather than one molecule at a time.

The federal US picture is the longest-running gap in the table. The FDAFood and Drug Administration announced a proposed rule to ban formaldehyde and formaldehyde-releasing chemicals from hair-smoothing and hair-straightening products in October 2023. The proposed rule has been repeatedly deferred — through 2024 and 2025 — and as of April 2026 no final rule has been published. The proposed rule would not, in any case, regulate DMDM hydantoin in shampoos and conditioners; it covers only hair-straightening formulations. The federal cosmetics regulator does not currently restrict any of the five major releasers in everyday personal-care products.

The releasers are not on the label as 'formaldehyde'. They appear under their own preservative names, usually toward the end of the ingredient list because they're used at fractions of a percent. The practical step is learning to spot them. Five names cover most of what you're likely to encounter.

An eso-friendly approach here is straightforward and cheap: spend ninety seconds reading the back panel before buying a shampoo or body wash. The five releaser names appear in plain English on the label of any product that contains them. The alternative products exist at the same price point. The system isn't asking you to do this; the system has had since 1988 to ask you to do this. You're doing it anyway.

DMDM hydantoin is a smaller story than the broader 'forever chemicals' picture covered in our PFAS profile and chemically simpler than the parabens family our parabens article walks through, but it's the cleanest illustration on the label of how the cosmetic preservative system actually works. The molecule does what it was designed to do — it stops the bottle going off. The mechanism for that stability is slow formaldehyde release. The chemistry has been on shampoo ingredient lists for forty years, and the labelling system treats the releasers as separate ingredients while the immune system treats them all as formaldehyde.

Washington's rule will phase out 28 releasers as a class in 2027. The EU has tightened the warning threshold by a factor of 50 — from 500 parts per million down to 10, low enough to capture every shampoo formulated with DMDM hydantoin at typical use concentrations. California regulates one of the five major releasers; the federal US regulator regulates none of them in everyday personal care. Five names on a back panel is the gap between what the system has decided you need to be told and what your skin is going to react to. Now you know the names.