What diapers are safe? Here’s Our Verdict

Our standard for diapers

Find & compare diapers

Ingredient concerns for diapers

• PFASAvoid entirelyAvoid entirely where you can identify it. This is not a "the dose makes the poison" ingredient in the way many flagged substances are: PFAS's defining feature is that it persists and accumulates, which is why regulators have moved toward restricting the entire chemical class rather than approving individual uses. The practical difficulty is that PFAS is rarely named directly on a label — it hides behind water-resistance or stain-resistance marketing claims, or isn't disclosed as a processing aid at all, which is why independent lab testing (not the label) is the only reliable way to know if a specific product contains it. A concrete example of what that testing looks like: a 2023 investigation had 65 diapers and diapering accessories from 40 brands tested by a U.S. EPA-certified lab for organic fluorine (a PFAS screening marker) and found detectable levels — 10 to 323 parts per million — in 23% of products overall (17% of disposables, 30% of cloth diapers and accessories), reviewed by four independent scientists including a former director of the National Institute of Environmental Health Sciences. Certifications that sound protective aren't a substitute for this kind of testing: OEKO-TEX Standard 100, for example, does not guarantee a product is PFAS-free, since its extractable-organic-fluorine method can miss fluoropolymers.
• Sodium polyacrylateGenerally fineGenerally fine. The Cosmetic Ingredient Review expert panel, which re-evaluates ingredient safety as new data emerges, has found sodium polyacrylate safe as used in personal-care and absorbent products, and it's classified as inert for EPA pesticide-formulation purposes at the low residual monomer levels manufacturing standards require. It's a reasonable ingredient to stop worrying about; the genuine diaper-safety conversation is elsewhere — in fragrance, lotion additives, and bleaching byproducts.
• Elemental chlorine-free pulpDepends on contextDepends on context, and mostly on which claim a label is actually making. Modern ECF pulp, as used by major diaper manufacturers, tests at dioxin levels regulators consider trace to non-detectable, and ECF is the pulp industry's overwhelming global standard today — it is not the same red flag as the elemental-chlorine process it replaced decades ago. The thing worth watching for is marketing language: a package that says "chlorine free" without specifying "totally" or "processed" almost always means ECF, which is a meaningfully different (and lower-risk) claim than TCF. It's a legitimate distinction to know, not a reason to treat every ECF product as equivalent to old-process bleaching. True TCF pulp remains a small share of total pulp production, which is exactly why the wording matters: if a brand doesn't explicitly say "TCF" or "Totally Chlorine Free," the safer assumption is ECF, not TCF — and it's worth double-checking that assumption periodically, since a manufacturer can switch a product from TCF to ECF pulp without necessarily updating its marketing copy to match.
• Zinc oxideUse with cautionUse with caution, and check the specific product before assuming "zinc oxide" alone tells you anything. Pharmaceutical- and cosmetic-grade zinc oxide from a reputable manufacturer can meet purity specifications (typically USP or equivalent) that cap heavy metal content at trace levels — but independent testing across 2025-2026 found detectable lead in nearly every zinc-oxide diaper-rash cream tested regardless of price or "clean" marketing, with several results exceeding Washington State's new cosmetic lead limit of 1 part per million (1,000 ppb, effective January 2025) by three to four times over. FDA's own cosmetic lead guidance — a non-binding 2016 draft recommendation, not an enforceable limit — puts a recommended maximum around 10 ppm (10,000 ppb), so these results sit well under federal guidance while still exceeding the stricter state standard; that gap is exactly why "not illegal" and "not detected" are different claims, and why a brand can legally sell a product that independent testing flags as concerning. One non-zinc alternative (a plain petrolatum-based balm, tested by the same initiative) came back non-detect for lead, cadmium, mercury, and arsenic — proof the contamination isn't inherent to treating diaper rash, only to this specific supply chain. See our diaper-rash-cream category page for product-by-product results.
• FragranceUse with cautionUse with caution, and treat "unscented" and "fragrance-free" as different claims. "Unscented" can mean a masking fragrance was added to cancel out a base odor — the fragrance mixture is still present. "Fragrance-free" means no fragrance ingredients were added at all, which is the claim that actually matters if avoiding undisclosed fragrance is the goal. For skin worn against skin for extended periods, especially on infants, fragrance-free is the more conservative and more verifiable choice — not because fragrance is always harmful, but because "always" isn't answerable without knowing what's in it.
• PhenoxyethanolUse with cautionUse with caution specifically in leave-on products applied to an infant's diaper area — wipes, barrier creams, and similar — where France's binding labeling requirement is most directly relevant, and where it's been in force since 2019, not a lapsed or superseded rule. The EU's own scientific committee disagrees and considers phenoxyethanol safe at up to 1% for all ages, so this isn't an "avoid everywhere" ingredient; it's a "check the product type, and know the regulators disagree" ingredient. A rinse-off product (shampoo, wash) is a different exposure than a leave-on wipe or cream used at every diaper change, and phenoxyethanol-free alternatives in both categories typically cost about the same.
• PhthalatesUse with cautionThis lands in caution rather than outright avoid, because the worst-evidenced phthalates are already restricted in the products where children's exposure would matter most. Since Section 108 of the 2008 Consumer Product Safety Improvement Act and a 2017 CPSC final rule (16 CFR Part 1307), DEHP, DBP, BBP, DINP, DIBP, DPENP, DHEXP, and DCHP have been permanently banned above 0.1% (1,000 ppm) in any accessible part of U.S. children's toys and child care articles; the EU's REACH regulation (Annex XVII, Entry 51, expanded and in force since July 2020 via Commission Regulation (EU) 2018/2005) caps DEHP, DBP, BBP, and DIBP at 0.1% in essentially all plasticized consumer articles, Entry 52 separately restricts DINP/DIDP/DNOP in mouthable toys and childcare items, and the EU Cosmetics Regulation bans several phthalates from cosmetics outright. A toy, teether, or child-care item bought new from a mainstream U.S. or EU retailer today is very unlikely to contain these chemicals at meaningful levels, and the FDA states it currently has no evidence that phthalates as used in cosmetics pose a safety risk. The real remaining exposure comes from three places: older, secondhand, or cheaply imported vinyl products made before or outside these rules; personal-care and fragranced products where the "fragrance" label can legally hide undisclosed phthalates like DEP; and general indoor dust from vinyl building materials. For pregnant people and infants specifically, it's worth actively minimizing exposure to DEHP, DBP, and BBP given the anti-androgenic evidence — favor fragrance-free products and non-PVC household goods where that's an easy swap — but for the general population this is a "reduce where convenient, not a reason to panic over any single labeled product" chemical family rather than a hard no.
• Formaldehyde-Releasing PreservativesUse with cautionThis lands as caution, not avoid — the Group 1 cancer classification that makes "formaldehyde" alarming applies to a different exposure route and dose (sustained occupational inhalation) than a rinse-off shampoo or leave-on lotion delivers, and both CIR and FDA continue to regard these preservatives as safe as used within their capped release limits for the general population. That's distinct from heat-activated hair-smoothing/straightening treatments, a different product category that uses formaldehyde or methylene glycol directly rather than as a slow-release preservative, and which FDA has separately warned can expose salon workers and clients to airborne formaldehyde. That said, regulators are visibly tightening the rules around this category, which is itself informative: the EU cut its cosmetics "releases formaldehyde" warning-label threshold fifty-fold, from 0.05% to 0.001% free formaldehyde, via Commission Regulation (EU) 2022/1181 of 8 July 2022 (phased in through 2026), and California's Toxic-Free Cosmetics Act (AB 2762) bans quaternium-15 outright, alongside formaldehyde, paraformaldehyde, and methylene glycol, in any cosmetic sold in the state as of January 2025 — though it does not ban every releaser in this family. The honest nuance: anyone with a known formaldehyde or quaternium-15 contact allergy should avoid this entire category, since cross-reactivity is common; it's also reasonable to prefer releaser-free formulas for leave-on products used on infants or on broken/irritated skin, where sensitization risk is highest. For ordinary rinse-off use by adults without a known allergy, this category doesn't belong in the same risk tier as an ingredient with direct, sustained exposure to Group 1 carcinogen levels — it's a real, worth-knowing allergen with a growing regulatory footprint, not a hidden carcinogen.
• Natural Rubber LatexDepends on contextFor the great majority of babies with no personal or family history of latex allergy, natural rubber latex in diaper elastic is not a meaningful worry: true Type I latex-protein allergy is uncommon in the general population, and sensitization typically builds up through repeated mucosal or medical exposure (surgical gloves, catheters, multiple procedures) rather than from occasional skin contact with a diaper's leg cuff. Notably, the documented cases of "diaper elastic" skin reactions in the dermatology literature — such as a 2020 Contact Dermatitis case report — trace back not to latex protein at all but to rubber-processing chemicals like mercaptobenzothiazole, producing a delayed Type IV allergic contact dermatitis; and both of these are still far rarer than ordinary diaper rash, which is overwhelmingly irritant contact dermatitis from prolonged urine/feces exposure and has nothing to do with elastic material. A true latex reaction in a diapered infant looks like localized contact urticaria — redness, itching, or hives right where the elastic touches skin, appearing within roughly 15 minutes and resolving after the diaper is removed — while systemic signs (widespread hives, swelling, wheezing, anaphylaxis) are rare with skin-only contact and are the ones that need urgent medical attention. For the minority who do carry real risk — a diagnosed latex allergy, latex-fruit syndrome, spina bifida, or a history of many early surgeries — avoiding natural rubber latex is medically sound, and a "latex-free" or "not made with natural rubber latex" diaper is a reasonable precaution, with the caveat that no label can guarantee zero trace protein. This is a genuine, well-characterized allergen worth flagging for the people it actually affects, not a reason for the average parent to treat every rash as a latex reaction.
• Dyes and ColorantsUse with cautionThis is a real, published mechanism rather than a theoretical worry, but it is also a genuinely uncommon one in absolute terms — the foundational case series involved only a handful of infants, and the great majority of babies wear colorfully printed diapers with no skin reaction at all, since the printed designs use surface pigments rather than the fiber-bound dyes implicated in the reports. It's also worth noting that several major U.S. diaper brands, including Pampers and Luvs, now publicly state they do not use disperse dyes in any part of their products — if accurate and representative of the broader current market, that would meaningfully shrink the specific pathway described below for those brands, though the underlying case series (2005) predates this apparent shift, didn't name brands, and it's unclear how uniformly the disclosure applies across store brands, cloth-diaper elastics, or older stock. For a baby with no rash history, colored diapers are not something to worry about. For a baby with a persistent, recurrent, or unusual diaper rash that isn't responding to standard diaper-rash care — especially one that follows the shape of the waistband or leg cuffs rather than the whole diaper area — dye-related allergic contact dermatitis is a documented possibility worth raising with a pediatrician, and switching to a plain-white, print-free, undyed-elastic diaper is a low-cost, evidence-informed elimination step rather than an overreaction.
• Volatile Organic Compounds (VOCs)Use with cautionRead the actual numbers before reacting to a "diapers contain carcinogens" headline - but read past the single reassuring number, too. The Michigan team's own worst-case modeling put the added lifetime cancer risk from some diapers at roughly 1 in a million, which sits at the more protective end of the cancer-risk range (1-in-1,000,000 to 1-in-10,000) EPA treats as acceptable when making cleanup decisions at contaminated sites. But that same study found a much less reassuring non-cancer number: in some diapers, VOC levels (driven largely by n-heptane) produced a hazard quotient as high as 1,609 under high-use exposure scenarios - hundreds of times above the level regulators generally treat as a concern threshold. The study's own authors concluded that risk from some diapers under high-use conditions (which isn't a rare or contrived scenario for an infant in diapers most of the day, or an adult incontinence user) is "high enough to warrant additional attention and possibly corrective measures." So this is a genuinely mixed picture, not a uniformly reassuring one: the cancer-risk math looks low, but the non-cancer VOC math, in at least some tested products under realistic heavy-use conditions, does not. A broader 2024 review in Science of The Total Environment, pooling this and other diaper-chemical exposure studies, concluded that non-cancer health risk to infants from diaper-related chemical exposure generally reads as acceptable under typical-use assumptions and is elevated mainly under extreme, worst-case assumptions - which is reassuring in aggregate but doesn't fully resolve the specific high-hazard-quotient products the 2023 study flagged. That's meaningfully different from the concern profile of PFAS, where persistence and bioaccumulation are the issue - trace VOCs off-gas and clear rather than build up in the body over time. The one finding actually worth acting on: "organic," "natural," and "sensitive skin" labeling showed no reliable correlation with lower VOC levels in the study, so don't pay a premium expecting that specific claim to solve this exposure - it doesn't. No US regulator currently sets a diaper- or hygiene-product-specific limit for benzene or total VOC off-gassing the way FDA guidance and some state rules cap 1,4-dioxane in rinse-off cosmetics, which is a real regulatory gap worth pressing manufacturers to close with published, independent VOC testing - including testing that reports non-cancer hazard quotients per product, not just aggregate cancer risk. Overall this reads as a "monitor it, ask brands for data, don't pay extra for an unverified 'organic' claim" level of concern for typical use, with a genuine open question about heavy, prolonged-wear scenarios - not a reason to avoid disposable diapers or hygiene products as a category.