Don’t take our word for it. Here’s every fact on this site, with its evidence grade, its caveats, and a link to the source, so you can check us.
We start you on the strong ones: findings that have been replicated or come from one large, well-controlled study. Moderate and weak facts are here too, labeled for what they are. Use “Link to this fact” to share a single card.
Children and pregnancy / Measured in people or real products
In 34 California childcare centers, the chemicals at the highest levels in the air children breathed came from cleaning and personal-care products: a silicone ingredient called D5 (median 51 µg/m³) and the citrus-scent chemical limonene (median 33 µg/m³).
Strong evidence: Single-day samples; upper values capped by calibration (underestimates).
Children and pregnancy / Review of other studies
About half of children's cosmetics in a Danish registry (634 of 1,179, 53.8%) listed perfume or a named fragrance allergen, and a US survey of 533 baby/child personal care products found a closely similar share (255/533, 47.8%; Bonchak 2018); in Danish baby-care products the commonest named allergens were linalool, limonene, and benzyl alcohol (the US primary does not report limonene or linalool by name).
Strong evidence.
Caveats
Danish half is primary (Botvid 2023): baby-care products specifically were 38.0% (178/468), the least fragranced category — do not write 'half of baby products' for Denmark. 'Fragranced' there includes products naming only an allergen such as benzyl alcohol (often a preservative). US half is now also primary (Bonchak 2018): 47.8% (255/533) across wipes/diaper care/soap/moisturizer/shampoo/sunscreen — but Bonchak's fragrance count is a ~20-keyword composite that does not report limonene or linalool individually, so the 'top three' allergen list is confirmed only for Denmark. Label counts remain lower bounds.
Source: Sukakul 2024, Acta Dermato-Venereologica Industry-funded Link to this fact
Children and pregnancy / Measured in people or real products
Pregnant women who had used perfume or cologne in the past day had 167% higher urine levels of MEP, the breakdown product of the fragrance solvent diethyl phthalate (DEP), than women who hadn't (177 women, Boston).
Strong evidence.
Caveats
Exposure marker, not health outcome; independently replicated in two further US cohorts with overlapping confidence intervals: Parlett 2013 (2.92x, postpartum, SFF multi-site) and Just 2010 (2.3x, 95% CI 1.6-3.3, pregnant, NYC minority cohort, with a dose-response and a null personal-air DEP association pointing to a dermal exposure route).
Source: Braun 2014, Journal of Exposure Science and Environmental Epidemiology Link to this fact
Children and pregnancy / Measured in people or real products
Synthetic musk fragrance chemicals were found in the breast milk of every one of 39 Massachusetts mothers tested in 2004; the most common, galaxolide (HHCB), was in 97% of them.
Strong evidence: One US state, 2004 samples, n=39; no product-use data; no health outcome.
Source: Reiner 2007, Environmental Science & Technology Link to this fact
Children and pregnancy / Measured in people or real products
Musk scent chemicals used in perfumes, deodorants, and detergents turn up in mothers' milk fat; a breastfed baby takes in an estimated ~2 micrograms of galaxolide a day.
Strong evidence.
Caveats
Intake is a model estimate from means of detects; the printed intakes do not reproduce from the stated method. Independently corroborated: Lignell 2008 (N=101, Swedish mothers, single lab) estimated median infant HHCB intake at 1650 ng/day (range 50-6950) using the same milk-intake modeling approach, closely matching this paper's printed 1830 ng/day figure in an unrelated country and cohort.
Source: Reiner 2007, Environmental Science & Technology Link to this fact
Children and pregnancy / Review of other studies
Indoor irritant limits are built on studies of healthy young adults; a 2024 toxicology review found only one controlled irritation study in children (none under age 5) and concluded no evidence-based safety factor for children could be set.
Strong evidence: Absence of data, not evidence of harm; the review lists reasons children could be more or less susceptible (e.g. more stable tear film).
Source: Kleinbeck & Wolkoff 2024, Archives of Toxicology Industry-funded Link to this fact
Children and pregnancy / Measured in people or real products
16.5% of children and adolescents patch tested in 9 population studies had a contact allergy, and in Swedish 16-year-olds fragrance (2.1%) was the second most common allergen after nickel.
Strong evidence: Any allergen, not fragrance only; some adolescent studies read too early and underestimate.
Children and pregnancy / Review of other studies
Children exposed to household chemical/cleaning products before birth or in their first year wheeze more often, now replicated in four independent prospective birth cohorts. The founding ALSPAC cohort (UK, n=7,019, Sherriff et al. 2005) found that an 11-product, frequency-weighted prenatal household chemical (TCB) score was significantly associated with persistent wheeze through 3.5 years (adjusted OR 1.06 per unit increase; top-vs-bottom-decile adjusted OR 2.3), but not with transient early or late-onset wheeze; this is the same ALSPAC cohort and composite-score methodology that both Casas 2013 and Mikeš 2019 name as their model. In the Canadian CHILD cohort (n=2,022), frequent use in early infancy was linked to higher odds of recurrent wheeze (adjusted OR 1.35) and asthma diagnosis (adjusted OR 1.37) by age 3, but not isolated skin-prick allergy (adjusted OR 1.14, ns), though wheeze combined with allergy was itself raised (adjusted OR 1.49). In the Czech ELSPAC-CZ cohort (n=3,411), each 1-SD increase in an 18-product household chemical exposure score at 6 months raised the odds of intermediate-onset transient, intermediate-onset persistent, and early-onset persistent wheeze phenotypes through age 5 (adjusted OR 1.23-1.36), and those persistent phenotypes carried 5-8x higher odds of doctor-diagnosed asthma at age 7 (pediatrician records, not self-report). In the pooled Spanish INMA cohorts (n=2,292), prenatal (third-trimester) use of cleaning sprays raised the odds of both wheezing (OR 1.37) and lower respiratory tract infection (OR 1.29) in the first year of life, with solvents raising wheezing odds (OR 1.30) and air fresheners raising LRTI odds (OR 1.29); bleach, ammonia, and non-spray cleaners showed no association.
Strong evidence.
Caveats
All four cohorts are observational and use self-reported product-use frequency, not measured indoor air chemistry. The ALSPAC/Sherriff 2005 and ELSPAC-CZ signals are specific to persistent-type wheeze phenotypes, not wheeze in general; ALSPAC found no association with transient early or late-onset wheeze. The 'not atopy' finding remains single-cohort (CHILD only); ELSPAC-CZ's atopic-rash variable is a covariate, not an outcome, and Casas 2013 measured neither atopy nor a physician asthma diagnosis, only maternal-reported LRTI and any-wheezing through 12-18 months, so it neither confirms nor refutes the atopy finding; ALSPAC/Sherriff 2005 did not assess atopy or asthma at all. Casas 2013 is the only cohort among the three post-2010 cohorts with genuinely prenatal (rather than early-postnatal) exposure measurement, but ALSPAC/Sherriff 2005 is also prenatal and its composite score behaved like CHILD's and ELSPAC-CZ's (significant), not like Casas's (null) — evidence against prenatal timing per se explaining the Casas null. Casas 2013's own summed cleaning-product-count score was NOT a significant predictor, unlike ALSPAC's, CHILD's, and ELSPAC-CZ's frequency-weighted composite scores — an open methodological tension (register entry C262) that means 'more products used' is not yet a uniformly replicated exposure metric across all cohorts, even though several specific products (sprays consistently, plus solvents and air fresheners) are.
Source: Salonen 2024, Environment International Link to this fact
Children and pregnancy / Review of other studies
The government scientists who first linked lavender and tea tree oils to breast growth in children agree that four case reports cannot prove the oils caused it, and have called for epidemiological studies.
Strong evidence: Correspondence; the same letter also calls ER-agonist exposure 'substantial evidence of causality'.
Source: Ramsey et al. 2020 (JCEM reply), The Journal of Clinical Endocrinology & Metabolism Link to this fact
Children and pregnancy / Review of other studies
A 2018 US EPA systematic review found robust evidence that two common phthalates, DEHP and DBP, are linked to male reproductive harm (lower sperm quality; shorter anogenital distance in baby boys exposed before birth) at everyday exposure levels.
Strong evidence: Associations graded by protocol; DEHP/DBP are mainly plasticizers, not fragrance ingredients; no meta-analysis.
Source: Radke 2018 (US EPA), Environment International Link to this fact
Children and pregnancy / Measured in people or real products
A perfume ('eau de cologne') marketed for babies listed 16 different fragrance allergens, the most of 1,179 children's products in a Danish registry.
Strong evidence: A single product (the maximum), not typical; mean 0.4-1.8 named allergens per product by category; label presence, not dose.
Children and pregnancy / Measured in people or real products
In a national Canadian birth cohort (n=2,022), infants in homes with more frequent use of household cleaning products at 3–4 months of age had higher odds of recurrent wheeze (adjusted OR 1.35 per interquartile increase in use) and of an asthma diagnosis (adjusted OR 1.37) by age 3, but no higher odds of isolated allergic sensitization (skin-prick positivity).
Strong evidence.
Caveats
Observational; the authors explicitly state their results do not prove causation. Recurrent wheeze combined with atopy was itself significantly increased (adjusted OR 1.49), so isolated atopy — not allergic sensitization in general — is what showed no association.
Source: Parks et al. 2020, CMAJ (Canadian Medical Association Journal) Link to this fact
Children and pregnancy / Measured in people or real products
Even products marketed 'fragrance free' for children commonly contain contact allergens: of 52 US pediatric products so labeled in 2013, 42 (80.8%) contained at least one NACDG-tray allergen and 29 (55.8%) contained two or more.
Strong evidence: Label-based; allergens are not broken down by which ones appeared specifically in the 'fragrance free' subgroup beyond the count, and no concentration data exist.
Source: Hamann 2015, Journal of Allergy and Clinical Immunology Link to this fact
Children and pregnancy / Government agency
Because of endocrine-disruption concerns, the EU has banned bisphenol A from baby bottles and other food containers for infants and young children, set very low BPA migration limits for other food-contact materials, and banned or restricted several parabens in cosmetics 'in particular to protect infants and young children.'
Strong evidence: These are specific, named EU regulations (bisphenol A: Reg. (EU) No 10/2011 and 2018/213; parabens: Reg. (EU) No 358/2014 and 1004/2014), not a general endocrine-disruptor rule.
Source: EU Commission 2018 Link to this fact
Children and pregnancy / Government agency
Benzophenone-3 is used in baby lotions, oils, and creams (not powder) at concentrations up to 0.25%, according to the cosmetics industry's own 2021 concentration-of-use survey.
Strong evidence: Figure is from the industry's own use-concentration survey (Personal Care Products Council), reported without an accompanying infant-specific risk assessment in this document.
Source: CIR 2021 Benzophenones Industry-funded Link to this fact
Children and pregnancy / Measured in people or real products
In a fully powered UK RCT of 1,394 high-risk infants, daily whole-body emollient from birth to age 1 did not prevent eczema at age 2 years (23% vs 25%, adjusted RR 0.95, 95% CI 0.78-1.16), despite 74-88% adherence and results consistent across eight different eczema definitions.
Strong evidence.
Caveats
This is a single large RCT, but it directly excluded the trial's prespecified 30% relative-risk reduction and was consistent across all secondary eczema definitions, so the null result itself is well supported; it does not address using emollient to treat established eczema.
Source: Chalmers et al. 2020 (BEEP trial), The Lancet Link to this fact
Children and pregnancy / Other evidence
FDA says it prioritizes enforcement against homeopathic products marketed to vulnerable populations including children, pregnant women, and the elderly, and against products for serious diseases such as cancer, non-oral/topical routes, injury reports, risky ingredients, and quality problems.
Strong evidence: Enforcement priorities, not a ban; the guidance is nonbinding and many products fall outside the six categories.
Source: FDA consumer page: Homeopathic Products, FDA web page Link to this fact
Children and pregnancy / Review of other studies
The same review rated most of the evidence on gestational diabetes, blood pressure, fetal growth and child neurodevelopment as low or very low quality (36 very low, 63 low, 7 moderate, 2 high across 108 graded rows), so those links are associations, not established harms.
Strong evidence: Count is the page author's tally of grades printed in the review's tables; the review's own headline is more confident than its grades.
Source: Almeida-Toledano 2024, Science of The Total Environment Link to this fact
Children and pregnancy / Review of other studies
A 2026 narrative review concludes that thyroid antibodies and moderate-to-severe iodine deficiency are the most consistent and reproducible predictors of thyroid dysfunction in pregnancy; women positive for both TPO and thyroglobulin antibodies had overt hypothyroidism in 7.0% versus 0.1% of antibody-negative women (OR 24.2).
Strong evidence: The review itself is ungraded narrative; the strength comes from the large pooled meta-analysis it cites.
Children and pregnancy / Other evidence
EFSA stresses that a higher hyperactivity score in the trial is not the same as ADHD: the score was not clinically validated, and the Panel says it is unknown whether the small changes would affect schoolwork.
Strong evidence: Panel's judgment; it also suggests a sensitive subgroup may exist whose size cannot be estimated.
Children and pregnancy / Other evidence
The EU's colors warning is a disclosure label that says an effect 'may' occur in children; it is not a ban on the six colors and does not change any acceptable daily intake.
Strong evidence: The Regulation's text gives no rationale; the link to the Southampton trial and EFSA's opinion comes from those source pages, not from this text.
Source: EU Reg 1333/2008 Link to this fact
Children and pregnancy / Other evidence
In a Swedish register study of 768,227 children (19,891 with ADHD), smoking in pregnancy went with about 2 times the ADHD risk (1-9 cigarettes/day HR 1.89; 10 or more HR 2.50), but the risk fell to 1.45 and 1.69 when cousins were compared and to 0.88 and 0.84 (not significant) when siblings were compared.
Strong evidence.
Caveats
One national register cohort (replicates smaller sibling studies in Finland, UK, US); ADHD ascertained from hyperkinetic diagnoses and medication, which likely captures severe cases; sibling designs are sensitive to non-shared confounders; tobacco smoke is not a fragrance exposure.
Source: Skoglund 2014, Journal of Child Psychology and Psychiatry Link to this fact
Children and pregnancy / Other evidence
In a Danish register study of about 968,000 children born 1991-2006, children whose mothers smoked in pregnancy had about twice the rate of ADHD (hazard ratio 2.01, 95% CI 1.94-2.07), rising to 2.77 for 10 or more cigarettes a day.
Strong evidence: Observational association in an ordinary cohort; the authors attribute it largely to family factors, not a proven effect of smoking.
Source: Obel 2016, Journal of Child Psychology and Psychiatry Link to this fact
Fragrance is everywhere / Measured in people or real products
In the US, everyday chemical products (paints, inks, glues, cleaners, personal care products, and pesticides) now release about twice as much air-polluting vapor as cars and trucks: 7.6 vs 3.5 million metric tons in 2012.
Strong evidence: All chemical products, not fragrance specifically; coatings/inks/adhesives are 58%.
Fragrance is everywhere / Measured in people or real products
Even in Perth, a city with very clean outdoor air, limonene and pinene (scent chemicals typical of citrus- and pine-scented cleaners and air fresheners) were found in 80-100% of homes when a sensitive test was used.
Strong evidence: Low concentrations (median ~1.5-2.4 µg/m³); sources not measured; 64 weekday-daytime visits.
Source: Maisey 2013, Atmospheric Environment Link to this fact
Fragrance is everywhere / Measured in people or real products
Americans spend about 69% of their time at home and only 7.6% outdoors, so small indoor sources close to people matter far more than their share of total emissions suggests.
Strong evidence: NHAPS time-activity survey.
Source: Weisel 2002, Environmental Health Perspectives Link to this fact
Fragrance is everywhere / Measured in people or real products
In EPA's large 1980s studies of US homes, limonene, the chemical behind lemon scents, was the airborne chemical with the highest average concentration in people's homes.
Strong evidence: 1980-87 TEAM data; no health risk estimated for limonene.
Source: Wallace 1991, Environmental Health Perspectives Link to this fact
Fragrance is everywhere / Government agency
The US Environmental Protection Agency lists air fresheners, cleaners, disinfectants, aerosol sprays, and cosmetics among the household products that release volatile organic compounds, both during use and while stored.
Strong evidence.
Source: US EPA 2026 Link to this fact
Fragrance is everywhere / Measured in people or real products
Scientists can identify which everyday products are polluting a city's outdoor air just by measuring a handful of specific chemical fingerprints: a silicone compound (D5-siloxane) for personal-care products and citrus/pine-scent chemicals (monoterpenes, mostly limonene) for fragranced products, confirmed by mobile-lab measurements across New York City, Pittsburgh, Chicago, and Denver.
Strong evidence: Outdoor air only; four US cities, no LA data; tracer status is based on population-density correlation and inventory apportionment, not a direct source test.
Source: Gkatzelis 2021, Environmental Science & Technology Link to this fact
Fragrance is everywhere / Measured in people or real products
In blood tests of 100 healthy young adults in Vienna, the synthetic musk galaxolide was found in 91% of samples and musk xylene in 79%; only 7 of the 100 people had no synthetic musks in their blood at all.
Strong evidence: Cross-sectional, healthy university students in one city; no health outcome measured.
Source: Hutter 2009, Science of the Total Environment Link to this fact
Fragrance is everywhere / Measured in people or real products
Two completely different ways of estimating pollution from chemical products, one based on direct air measurements and one based on how much of each product gets sold and used, agreed with each other to within about a quarter.
Strong evidence: Agreement within 25% for the bulk comparison and within 30% for more than half of individual compounds compared, in Boulder and NYC only.
Source: Gkatzelis 2021b, Environmental Science & Technology Link to this fact
Fragrance is everywhere / Measured in people or real products
Musk fragrance chemicals were found in the breast milk of all 101 Swedish first-time mothers tested; the musk chemical galaxolide (HHCB) was detected in every single sample across a seven-year span (1996-2003).
Strong evidence: One Swedish county, single lab/method, N=101; no infant health outcome measured.
Source: Lignell 2008, Environmental Science & Technology Link to this fact
Fragrance is everywhere / Review of other studies
Human skin releases more than 400 different volatile compounds in total, but any single sampling study of airborne skin volatiles only ever captures 20 to 90 of them, so the reported 'chemical signature' of human skin varies hugely from paper to paper depending on the method used, not necessarily on real biological differences.
Strong evidence.
Caveats
Synthesis across 24 sampling studies and a 31-reference compilation of detected compounds (this review's own Tables 1 and 2), not a single new measurement; no head-to-head method comparison exists in the underlying literature.
Source: Dormont et al. 2013, Journal of Chemical Ecology Link to this fact
Fragrance is everywhere / Government agency
There are no longer any pristine, unpolluted places on Earth: endocrine-disrupting chemicals such as PFOS have been measured in marine-mammal tissue from the Arctic to remote Southern Hemisphere waters, far from any point source, carried by long-range air and ocean transport and biomagnified through food webs.
Strong evidence: Illustrated with wildlife biomonitoring data (marine mammals), not a direct measurement of fragrance or consumer-product chemicals specifically.
Source: WHO/UNEP 2012 Link to this fact
Fragrance is everywhere / Review of other studies
UV filters from sunscreens have been detected in marine water, sediment, and biota worldwide, from touristy coastal waters and coral reefs to remote regions with no direct sunscreen use, including Antarctica and the Arctic, at concentrations from nanograms to milligrams per liter.
Strong evidence.
Caveats
Ubiquity claim rests on many independent detection studies across regions, not one measurement; the concentration range spans several orders of magnitude and is not broken down per compound in every location cited.
Source: Hodge 2025, Marine Pollution Bulletin Link to this fact
Fragrance is everywhere / Measured in people or real products
General-population blood styrene levels (US NHANES, general adult population) run in the range of roughly 0.02-0.2 nanograms per milliliter (50th to 95th percentile across several 1988-2008 survey cycles) — two to three orders of magnitude below levels reported in occupationally exposed reinforced-plastics or rubber workers.
Strong evidence: NHANES biomonitoring data; does not by itself establish health risk at these levels, only relative exposure magnitude versus occupational cohorts.
Source: IARC 2019 (Vol. 121) Link to this fact
Fragrance is everywhere / Government agency
Benzophenone-3 (oxybenzone) is used in fragrance formulations at concentrations up to 0.5%, and separately in lipstick (up to 0.5%, incidental ingestion) and bath soaps/detergents (up to 0.5%, mucous-membrane contact), per the cosmetics industry's own 2021 concentration-of-use data.
Strong evidence.
Caveats
Figures are from the industry's own use-concentration survey; the fragrance-formulation use is not separately broken out by product category in the public report beyond this single statement.
Source: CIR 2021 Benzophenones Industry-funded Link to this fact
Fragrance is everywhere / Government agency
Canada's own 2026 fragrance-allergen disclosure mandate directly incorporates this EU Annex III list by reference, so the EU's count discrepancies (24 vs 26 originals, 56 additional, ~80 total) propagate directly into Canadian cosmetics law as well.
Strong evidence.
Caveats
Confirmed via Health Canada's own guidance (src-2026-health-canada-cosmetic-labeling-fragrance-disclosure), which cites '24 to 81' — closely matching this regulation's own 24+56=80 arithmetic.
Source: Regulation (EU) 2023/1545 Link to this fact
Fragrance is everywhere / Journalism
Skin Deep's 1-10 hazard score is relative to EWG's own database: the most hazardous 5 percent of products and ingredients are all set to 10 and the rest scaled down, so a 10 does not mean a fixed level of danger.
Strong evidence: Method as published by EWG. The page gives no cut-offs for its "low, moderate, high" bands; the common 1-2, 3-6, 7-10 grouping is from Barrett 2025, not this page.
Source: EWG Skin Deep methodology page (2026 capture) Link to this fact
Green, natural and unscented / Measured in people or real products
'Natural' is no protection: lavender oil's own components form skin allergens when exposed to air.
Strong evidence.
Caveats
Primary is Hagvall 2008 (lab open-flask air exposure, 10-45 weeks): the oil oxidized at the same rate as a synthetic mix of its terpenes; 3.3% linalyl acetate-OOH + 0.48% Lin-OOH at 10 weeks; mouse LLNA EC3 36% fresh -> 4.4% at 45 weeks (weak -> moderate); 4/4 oxidized-linalool-allergic patients reacted to oxidized oil (elicitation only, no fresh-oil control). Real-bottle oxidation rates not measured. Lavender oil ~40% linalool (Klaschka 2016).
Source: Bråred Christensson 2016, Contact Dermatitis Industry-funded Link to this fact
Green, natural and unscented / Review of other studies
Fragrance allergens turn up in "fragrance-free" moisturizers and personal-care products across at least four independent surveys: Scheinman named ten US "fragrance-free" products with fragrance ingredients in 1999 (benzyl alcohol in 5 of 10), dermatologists named nine more major-brand examples containing benzyl alcohol in 2007 (including a baby lotion), two of Scheinman's own named products (Cetaphil Moisturizing Cream, Moisturel Therapeutic Lotion) reappear on this list, a 2016 survey of best-selling US body moisturizers found 18 of 40 labeled "fragrance free" (45%) listed at least one ingredient dermatologists flag for fragrance-allergic patients, and a 2015 survey of 187 US pediatric personal-care products found that of the 52 labeled "fragrance free," 42 (80.8%) contained at least one contact allergen from the standard NACDG screening tray (55.8% had >=2).
Strong evidence.
Caveats
Four independently compiled surveys (1999, 2007, 2016, 2015) across adult and pediatric products now agree that 'fragrance-free'-labeled products commonly contain contact allergens. Hamann 2015's screening tray is broader than fragrance chemicals alone (also preservatives, surfactants; preservatives were in fact its largest allergen category), so its 80.8% is not a pure fragrance-allergen count like Xu 2017's CAMP cross-reactor tally, but the direction and scale match. Still label-based, not chemical analysis; don't name current products without checking today's labels.
Green, natural and unscented / Review of other studies
In the US, "fragrance free", "unscented" and "free and clear" have no legal definition, so each company picks its own meaning; some "unscented" products contain a masking fragrance added to hide the product's smell.
Strong evidence.
Caveats
Expert review; the legal-definition point is not itself quantified, but the practical consequence now is: Hamann 2015 found 42 of 52 (80.8%) US pediatric products labeled 'fragrance free' contained >=1 NACDG-tray contact allergen, converging with Xu 2017's 45% (18/40) for adult moisturizers (narrower CAMP cross-reactor definition). Two independent, systematically screened surveys across different age populations now confirm the practical failure of the 'fragrance free' label; masking fragrance specifically remains unmeasured (see Scheinman 1999 for the one documented case). Strong because the absence of a legal definition is a regulatory fact. Hamann's 80.8% counts any NACDG allergen (mostly preservatives), not fragrance; use Xu 2017's 45% for fragrance-related allergens.
Green, natural and unscented / Industry source
Even the fragrance industry's safety institute said natural fragrance is not safer: "There is no link between the natural origin of fragrance materials and their safety."
Strong evidence: An opinion with no data in the chapter; it agrees with independent emission tests of 'green' products.
Source: Ford 1994 (RIFM) Industry-funded Link to this fact
Green, natural and unscented / Government agency
Health Canada's current guidance (accessed 2026-09-27) states that a cosmetic labeled 'unscented' or 'fragrance-free' may contain a masking agent added to hide other odors, confirming a 2010 secondhand quote of the same policy word for word.
Strong evidence.
Caveats
Confirmed directly against Health Canada's current (2025-03-17/2026-04-24) guidance and its Cosmetic Regulations s.21.4(3) legal basis; no longer resting on a 2010 secondhand NGO quote alone. Does not measure how often masking actually occurs in Canadian products.
Green, natural and unscented / Measured in people or real products
'Natural' oils get no special protection from air: in lab tests, lavender oil's main scent chemicals oxidized just as fast as a synthetic copy made only of those chemicals, and formed the same allergenic hydroperoxides.
Strong evidence: Lab chemistry, neat oil in open flasks over 10-45 weeks; one oil type (lavender), two batches.
Source: Hagvall 2008, Contact Dermatitis Industry-funded Link to this fact
Green, natural and unscented / Government agency
FDA states that products labeled "unscented" may still contain fragrance ingredients, because a manufacturer may add just enough fragrance to mask the smell of other ingredients without giving the product a noticeable scent.
Strong evidence: FDA gives no formal regulatory definition of "unscented" or "fragrance-free" anywhere on the page; this is the only statement it makes about either term.
Source: FDA 2026 (Fragrances in Cosmetics) Link to this fact
Green, natural and unscented / Government agency
In Canada, a cosmetic labeled 'unscented' or 'fragrance-free' can legally still contain a fragrance ingredient added specifically to mask another smell, declared only as 'parfum' or 'fragrance' on the ingredient list.
Strong evidence.
Caveats
Health Canada's own current guidance (accessed 2026-09-27) and the Cosmetic Regulations s.21.4(3) legal basis; confirms and dates a 2010 secondhand quote of the same policy (David Suzuki Foundation 2010). It does not measure how often masking actually occurs in Canadian products; no Canadian product-testing study of this specific question is in our evidence base.
Source: Health Canada 2026 Link to this fact
Green, natural and unscented / Other evidence
The National Eczema Association's Seal of Acceptance requires products to be fully fragrance-free — "there should be no discernable/strong scent or anything aromatic in nature in the final product" — with no stated exemption for masking fragrance, and NEA directly checks a physical product sample for absence of scent as part of the review.
Strong evidence.
Caveats
This describes NEA's written rule and process, not a chemical measurement of any Seal-holding product; a more detailed 'clinical testing requirements' PDF is referenced but was not captured on this page.
Source: NEA 2026 Link to this fact
Green, natural and unscented / Other evidence
Seal of Acceptance products must also exclude a named 34-ingredient Ecz-clusion List (including cinnamic alcohol, cinnamic aldehyde, coumarin, eugenol, farnesol, geraniol, hydroxycitronellal, isoeugenol, oak moss absolute, and several citrus extracts), chemical UV absorbers other than titanium/zinc/iron physical sunscreens, and formaldehyde releasers, in any amount.
Strong evidence: Ingredient-list exclusion, verified by NEA's own review, not independent chemical assay of finished products.
Source: NEA 2026 Link to this fact
Green, natural and unscented / Other evidence
EWG Verified does not mean fragrance-free: fragranced cleaners can be licensed with fragrance up to 1 percent of the formula.
Strong evidence: Pet grooming products are limited to an EWG 'acceptable fragrance list' not reproduced in the document.
Source: EWG Verified criteria for cleaning products (10.24), Environmental Working Group, 2024 Link to this fact
Green, natural and unscented / Other evidence
OEKO-TEX Standard 100's limit table has no limit and no list for fragrance chemicals or fragrance allergens; the only fragrance-related rule is an odor test that requires 'no odor from mold, high boiling fraction of petrol, fish, aromatic hydrocarbons, or perfume'.
Strong evidence: Edition 02.2025 table only; the 2026 change notice mentions no fragrance row but the full 2026 table was not captured; odor test is a sensory judgment with no stated scale.
Source: OEKO-TEX Service GmbH, Standard 100 limit values Annexes 4-5, Ed. 02.2025 Industry-funded Link to this fact
Health effects / Animal study
Some older fragrance ingredients were neurotoxic: the nitromusk musk ambrette caused nerve damage in animal studies and was later withdrawn.
Strong evidence.
Caveats
Primary now read directly: Spencer 1984 is a large (N=220), dose-ranging, dual-route (dietary+dermal), histopathologically confirmed (light+EM) study showing clear dose-response demyelination and distal axonal degeneration. Note: musk ambrette's actual withdrawal from skin-contact products was driven primarily by a separate photoallergy finding (Cronin 1984), not directly by this neurotoxicity data, which RIFM had already argued had an adequate safety margin (>600) via absorption data (Ford 1994). Doses producing consistent neurotoxic effects are 50-250x the industry's own estimated maximum human exposure.
Source: Curtis 2004 (citing Spencer 1984), Environmental Health Perspectives Link to this fact
Health effects / Review of other studies
Chamber studies of asthma exclude moderate and severe asthmatics by design, and the one study that directly compared severity bands found the excluded group had the largest airway response, so these studies likely underestimate how sensitive people with asthma really are.
Strong evidence.
Caveats
Confirmed by a separate 103-study systematic review (Johansson 2016), whose own quality criteria excluded moderate/severe asthma in every included study; its one direct severity comparison (Linn et al. 1987, sulfur dioxide) found the moderate/severe group had the largest SRaw/FEV1 impairment, though not statistically significant in that small sample. Direction is now doubly confirmed (two independent reviews built on the same design flaw); magnitude of the underestimate is still not quantified.
Source: Kleinbeck & Wolkoff 2024, Archives of Toxicology Industry-funded Link to this fact
Health effects / Review of other studies
Chamber studies show irritant gases cause severe airway narrowing in people with asthma at concentrations at or below standard workplace exposure limits: asthmatics had severe bronchoconstriction to sulfur dioxide at about 0.4 ppm (1.1 mg/m3) after just 5-10 minutes of exercise, and to sulfuric acid at about 1 mg/m3.
Strong evidence.
Caveats
Quantified by Johansson 2016 (103-study systematic review): asthmatics showed severe airway narrowing (SRaw >=200%, FEV1 drop >=20%) to sulfur dioxide at >=1.1 mg/m3 (~0.4 ppm) after 5-10 min exercise and moderate-severe responses to sulfuric acid around 1 mg/m3 (oronasal) or 0.35 mg/m3 (mouth-only); both sit at or below typical occupational exposure limits for these substances (e.g. ACGIH SO2 TWA ~2 ppm/STEL 5 ppm; background knowledge, verify against current ACGIH tables). This is now primary chamber data, not only a secondary citation of Baur 2012's occupational-asthma review. Still not specific to fragrance chemicals; these are combustion/industrial irritant gases. Strong for sulfur dioxide and sulfuric acid only; no equivalent chamber data exist for fragrance chemicals, so do not transfer the grade to fragrance.
Source: Kleinbeck & Wolkoff 2024, Archives of Toxicology Industry-funded Link to this fact
Health effects / Review of other studies
People adapt to a steady odor, so a constant background scent stops being noticed while it is still in the air.
Strong evidence.
Caveats
Adaptation reduces perceived odor, not exposure; well established in chemosensory science but cited here from a review. Applies to a typical/nonsensitive population: Andersson et al. 2009 found nonsensitive controls habituated to CO2 chemosensory exposure over ~1.5 h as this claim describes (perceived intensity decreased over time, P=.045), but chemically sensitive individuals (CSS-SHR-defined) did NOT habituate to the same stimulus (P=.322, ns) and instead showed harder-to-disengage attention (shorter chemosensory/olfactory P2 ERP latencies in the ignore condition, P=.003/.017) — so 'stops being noticed' does not hold for sensitized/chemically sensitive individuals, the population most relevant to fragrance-sensitivity consumer claims. Dalton 2000's psychophysics review (the primary literature this 2010 review draws on for the mechanism) backs the general-population claim with independent replication across decades and species: Wuttke 1999 (Drosophila), Stone et al. 1972 and Cain 1969-1977 (human threshold and suprathreshold adaptation), Stuiver 1958 (unilateral nasal adaptation), de Wijk 1989 (adaptation reaching apparent total disappearance), and a home-exposure study Dalton (2000) cites as unpublished data (n=10, benzyl acetate/geraniol) showing 2 weeks of continuous home exposure elevates detection thresholds for the specific odorant, with recovery to baseline taking roughly 3 weeks for most subjects. CORRECTION (2026-10-01): that n=10/benzyl acetate/geraniol study is NOT the formally published Dalton & Wysocki (1996, Percept Psychophys 58(5):781-792) paper — read directly, the published paper is n=8 (citralva vs. iso bornyl acetate), with recovery tracked for 2 weeks post-removal (half of subjects not fully recovered, one not yet begun; see src-1996-nature-duration-adaptation-long-term-odor-exposure and C367, now resolved). The published paper independently corroborates the same general-population mechanism at a smaller n, so the 'strong' rating is unaffected; only the description of which dataset is which has been corrected. This breadth of independent replication (now including a directly-read primary, not just secondhand reviews) meets our 'strong' bar for the general-population mechanism; the sensitized-population caveat above is unchanged and still applies.
Source: Dalton & Jaén 2010, Current Opinion in Allergy and Clinical Immunology Industry-funded Link to this fact
Health effects / Review of other studies
In the same EPA review, diethyl phthalate (DEP), the phthalate used in fragrance, had only 'slight' evidence of male reproductive effects, the weakest of six phthalates, even though people are exposed to it the most.
Strong evidence: Use to set the record straight, not to accuse fragrance; 'slight' is not 'none' (more high-confidence studies needed).
Source: Radke 2018 (US EPA), Environment International Link to this fact
Health effects / Measured in people or real products
In the largest controlled test of its kind, 130 healthy young women breathing an indoor chemical mixture with and without ozone for 140 minutes showed no significant change in symptoms, lung function, attention, or the stress hormone cortisol.
Strong evidence: Healthy women only; asthma and chemical sensitivity excluded; one 140-min exposure at 1.8 air changes/h; eye-blink rate not measured.
Source: Fiedler 2005, Environmental Health Perspectives Link to this fact
Health effects / Measured in people or real products
A systematic review of 103 controlled chamber studies found asthmatics develop severe airway narrowing to sulfur dioxide at concentrations around 1 mg/m3 (about 0.2-0.4 ppm) after just 5-10 minutes of exercise, while healthy people showed no response even at 15 mg/m3; individual dose-response modeling put asthmatics about 9 times more sensitive to sulfur dioxide than healthy people.
Strong evidence: Sulfur dioxide is a combustion/industrial gas, not a fragrance chemical; asthmatic subjects were mild cases only (severe asthma excluded by design).
Source: Johansson 2016, Critical Reviews in Toxicology Link to this fact
Health effects / Measured in people or real products
Every controlled chamber study in a 103-study review of asthma and airborne chemicals excluded people with moderate or severe asthma by design; the one study that directly compared severity groups found the more severe group had the largest airway response, though not statistically significant in that small sample.
Strong evidence: Direction, not magnitude: the comparison (Linn et al. 1987) was underpowered and did not reach significance.
Source: Johansson 2016, Critical Reviews in Toxicology Link to this fact
Health effects / Animal study
Musk ambrette, once a widely used musk fixative in fragrances (over 100,000 lb/year in the US since the 1920s), caused nerve-fiber demyelination and degeneration in a large industry-funded rat study of 220 animals, with hindlimb weakness ranging from isolated cases at moderate doses to complete weakness at the highest dose tested.
Strong evidence.
Caveats
Large (N=220), dose-ranging, dual-route (dietary+dermal) primary with clear dose-response and light+EM histopathology confirmation; doses producing consistent effects are 50-250x the industry's own estimated maximum human exposure.
Source: Spencer 1984, Toxicology and Applied Pharmacology Industry-funded Link to this fact
Information, not medical advice. See also: myths we won’t tell you.