Chemicals in Toothpaste: What Is in It and How to Choose a Safer One

Brush twice a day and that's 730 brushings two a day for one year — and around 51,100 across an average lifetime a year — and a measurable fraction stays with you every time. A 2023 systematic review in Frontiers in Public Health found that across studies of toothpaste ingestion, the proportion of paste swallowed ranged from 10.7% to 80.5% across the studies in the Petrović 2023 systematic review of toothpaste ingestion depending on age and brushing technique Petrović et al. 2023. Children swallow more. Adults swallow less, but still measurably. The rest passes through the oral mucosa, where it has at least a few minutes to do whatever it does. Part of our oral-care chemicals guide covers the broader picture; this article goes through the eight ingredients you actually need to recognise.

Beyond the fluoride or hydroxyapatite doing the cleaning, the rest is mostly formulation chemistry. The eight ingredients below come up repeatedly in regulatory filings, dental-research journals and consumer testing. Some are clear concerns with multiple human studies. Some are absurd-regulatory-status problems — chemicals banned in food but legal in mouths, banned from soap but legal between teeth. One is a targeted concern that affects roughly one in ten people. The point of the list isn't to make any of them sound dramatic in isolation. It's to give you the names so you can read the back of the tube and recognise what you're looking at. An eso-friendly tube is mostly a tube without these — or with the smallest defensible subset of them.

Start with the foam. Two-thirds of commercial toothpastes foam because of sodium lauryl sulfate — a 2024 survey of 160 varieties across 19 brands found SLSSodium lauryl sulfate — an anionic surfactant that creates foam in toothpastes, shampoos and detergents. CAS 151-21-3 in 67% of 160 commercial toothpaste varieties surveyed (Ucuncu 2024) — the highest-reported toothpaste-specific prevalence figure in the recent dental literature of products tested Ucuncu et al. 2024. Brushing and the active ingredient are what actually clean your teeth. The foam is cosmetic — it spreads the paste around the mouth and signals to your brain that something is happening. That's the entire mechanism.

SLS is an irritant. It's not an endocrine disruptor; the SLS chemical profile walks through why five separate cancer agencies have looked at it and found nothing. The toothpaste-specific question is narrower: SLS denatures mucinthe glycoprotein layer that lubricates and protects the mucosal lining inside your mouth, the protective glycoprotein layer covering the inside of your gums, cheeks and tongue. With repeated exposure, that layer thins. In a population that's already prone to RASRecurrent aphthous stomatitis — recurrent painful oral ulcers, affecting roughly one in five people across their lifetime, that thinning matters.

The cleanest experiment on this question is a Norwegian crossover trial from 1996. Thirty patients with active recurrent aphthous ulcers used three different toothpastes for six weeks each, in blinded order: one with sodium lauryl sulfate, one with cocoamidopropyl betaine — a milder amphoteric surfactant — and one with no surfactant at all. The dental team at the University of Oslo counted ulcers at every visit. Ulcer frequency was significantly higher on the SLS paste than on either alternative Herlofson & Barkvoll 1996. The effect was obvious on a sample of thirty. The trial doesn't quote a specific percentage for the magnitude — but the direction has been replicated in subsequent SLS-vs-CAPB studies, and the clinical recommendation has been steady for thirty years: if you get recurrent mouth ulcers, switch off SLS toothpaste before reaching for a steroid mouthwash.

The evidence doesn't support broad SLS avoidance. If it isn't bothering you, the energy you'd spend hunting for SLS-free toothpaste is better spent on the chemicals later in this list. The SLS-free formulations that do exist — most natural-brand toothpastes, plus single-ingredient bars and tablets — substitute cocamidopropyl betaine, sodium cocoyl isethionate, or decyl glucoside, all well-tolerated alternatives. The rest of the formulation is the same.

Triclosan was the headline antibacterial in Colgate Total — for two decades the world's most-used triclosan toothpaste — and the regulatory story around how it left the soap aisle but stayed in the dental aisle is still the cleanest example of how toothpaste is treated differently from everything else on the same shelf. Aiello, Larson and Levy reviewed the controlled trials of triclosan-containing consumer hand soaps and found that at the 0.1–0.45% wt/volthe consumer-product concentration range reviewed in Aiello et al. 2007 — covering most retail antibacterial soaps and washes concentrations used in retail products, triclosan washes performed no better than plain soap at preventing illness or reducing hand bacteria Aiello et al. 2007. The FDA cited that paper as the foundation of the 2016 consumer antiseptic-wash final rule.

The toothpaste evidence was different. Riley and Lamont's 2013 Cochrane review pooled thirty trials of triclosan-fluoride toothpaste against fluoride-only toothpaste — 14,835 participants in total — and found a small but consistent benefit. Plaque scores fell by a mean difference of 0.47 on a 0–5 scale at six to seven months (moderate-quality evidence); gingivitis fell by 0.27 on a 0–3 scale at six to nine months (moderate-quality); coronal caries reduced by 0.16 D(M)FS over two to three years (high-quality); root caries by 0.31 (moderate-quality) Riley & Lamont 2013. The benefit was real. It was also small.

The FDA's antiseptic-wash rule landed on 6 September 2016 (codified at 81 FR 61106), banning nineteen active ingredients — triclosan and triclocarban among them — from over-the-counter consumer hand and body washes, with a one-year compliance window before it took effect on 6 September 2017. The rule covered OTC monograph antiseptic products. Colgate Total wasn't one of them. Colgate had filed Colgate Total under NDANew Drug Application — the FDA pathway for a drug-marketing approval reviewed individually rather than under a class monograph 020231 in 1997, with triclosan as the gingivitis-active ingredient — making it the only triclosan toothpaste with a stand-alone FDA new-drug approval. One company's NDA kept a banned chemical on shelves for more than a year after the soap version was pulled.

Colgate reformulated to Colgate Total SF — stannous fluoride, no triclosan — in early 2019. The reformulation was voluntary. The EU side runs in parallel: Commission Regulation (EU) No 358/2014, adopted on 9 April 2014, capped triclosan at 0.3% the maximum permitted concentration of triclosan in toothpaste under EU Regulation 358/2014 (still in force as of 2026) in toothpaste, hand soap, body soap, non-spray deodorants, face powders, blemish concealers and nail-cleaning products. That cap is still in force. Triclosan is therefore legal in EU and UK toothpaste at up to 0.3% — but no major brand still uses it. The full triclosan profile covers the thyroid and antibiotic-resistance concerns separately.

Titanium dioxide is the reason your toothpaste is white. It has no therapeutic function. It doesn't clean teeth. It doesn't strengthen enamel. It's a white pigment listed on toothpaste tubes as CI 77891Color Index 77891 — the cosmetics-regulation name for titanium dioxide as a pigment, and its job is to make the paste squeezed onto your brush look like toothpaste rather than the translucent gel it would otherwise be. A 2020 PLOS ONE survey of 401 households in Madison, Wisconsin found titanium dioxide in 72.2% of US toothpastes — the largest-published prevalence figure for titanium dioxide in toothpaste, from a 401-household survey of personal-care products of the toothpastes used by participating families Wu et al. 2020. The Ucuncu 2024 Turkish survey put it at 68% of products. It's the rule, not the exception.

Here's where it gets strange. On 14 January 2022, the European Commission adopted Regulation (EU) 2022/63. Just under seven months later, on 7 August 2022, titanium dioxide stopped being a permitted food additive in the EU. The basis was EFSAEuropean Food Safety Authority — the EU agency that evaluates risks associated with the food chain's 2021 re-evaluation, which concluded that 'a concern for genotoxicity could not be ruled out' and that E171 'can no longer be considered as safe when used as a food additive' Younes et al. (EFSA FAF Panel) 2021. EFSA couldn't establish an acceptable daily intake — there was no longer a dose at which it could declare the additive safe. The same regulator still permits titanium dioxide as a colourant in toothpaste under Annex IV of Cosmetics Regulation 1223/2009. Banned from your croissant. Still in your mouth. Same regulator.

An earlier rat-feeding study made the same point on the gut side. Orally administered E171 accumulated in Peyer's patchesspecialised lymphoid tissue lining the small intestine — the gut's main interface with food antigens and the immune system after one week of dosing and significantly increased aberrant crypt foci in the colon — preneoplastic lesions — over a hundred days at 10 mg/kg body weight per day Bettini et al. 2017. That paper sat in EFSA's bibliography during the 2021 re-evaluation. The toothpaste exposure isn't the same as the food exposure that prompted the EU ban — most of a brushing is rinsed out — but the regulatory split goes one way only: food banned, cosmetics still allowed. If genotoxicity is the reason food-grade titanium dioxide is no longer considered safe at any dose, the cosmetics carve-out is a regulatory question of timing, not science.

Fluoride is the contested topic in oral care, and the answer depends entirely on whether the fluoride is on your teeth or in your bloodstream. Topical fluoride at consumer-toothpaste concentrations of 1000–1500 ppm has high-certainty evidence for preventing tooth decay. The 2019 Cochrane network meta-analysis of 96 trials put the prevented fraction at 23% 95% credible interval 19–27% — the prevented fraction for fluoride toothpaste at 1000–1250 ppm versus placebo, in children and adolescents (Walsh 2019 Cochrane) (95% credible interval 19–27%) for fluoride toothpaste at 1000–1250 ppm versus placebo, in children and adolescents, with a high-certainty rating for the comparison Walsh et al. 2019. That part isn't in serious dispute.

The contested part is what happens when fluoride is swallowed and reaches the developing brain. The cleanest evidence is from a Mexico City birth cohort. Bashash and colleagues enrolled pregnant women into the ELEMENTEarly Life Exposures in Mexico to Environmental Toxicants — the Mexico City prospective birth cohort that tracked maternal fluoride exposure and child cognition cohort, measured fluoride in maternal urine during pregnancy as a direct biomarker of circulating exposure, then waited until the children were old enough to test. Among 211 mother-child pairs followed to ages six to twelve, every 0.5 mg/L increase in maternal urinary fluoride was associated with a 2.50 IQ point decrease per 0.5 mg/L increase in maternal urinary fluoride at ages 6–12 (Bashash 2017) IQ point decrease in the child (95% CI −4.12 to −0.59) Bashash et al. 2017. Mexico City's water isn't contaminated; the exposure is fluoridated salt and naturally fluoride-containing groundwater at levels comparable to North American averages.

A Canadian replication (Green et al. 2019, six cities, n=400 by intake / n=512 by urinary biomarker) reported a comparable signal with a sex interaction concentrating the effect in boys. The 2024 NTPNational Toxicology Program — the US federal program that conducts toxicology assessments for the public systematic review concluded with moderate confidence that fluoride exposure above 1.5 mg/L in drinking water is associated with lower IQ in children. Five weeks later, a federal judge in San Francisco ruled that the 0.7 mg/L US fluoridation target poses an unreasonable risk of reduced IQ in children under TSCA §6 and ordered EPA to initiate rulemaking (2024, Food & Water Watch v. EPA, Case 17-cv-02162-EMC). The full fluoride profile walks through the cohort findings, the NTP review and the court ruling in depth.

What this means for toothpaste: the systemic-versus-topical distinction is real and load-bearing. The fluoride in toothpaste is meant to act on enamel and then be spat out. Adults swallow a small fraction; small children swallow more — and dental fluorosisenamel mottling caused by excess fluoride exposure during tooth development, mostly cosmetic at mild levels is a separate, well-established reason to limit them to pea-sized doses. For adults, the dental benefit at 1000–1500 ppm is robust enough that switching to fluoride-free toothpaste trades a known caries-prevention effect for a hypothetical neurodevelopmental one. For pregnancy and early childhood, the calculation is different. The maternal-urinary findings are about systemic exposure during gestation; if you're pregnant and you also live in a fluoridated water area, fluoride-free toothpaste is one of the few exposure-reduction levers within easy reach.

Then the rest of the back panel. Four more ingredients show up often enough to recognise — even if the toothpaste-specific exposure is smaller and the evidence thinner than the four already covered.

Propylene glycol sits in oral-care formulations at 0.5–2% by weight, well below the 5–20% cosmetic ranges used in moisturisers and shampoos — the lower end is where it stays clear of the allergic-contact-dermatitis threshold. It's a humectant: keeps the paste moist so it doesn't dry out in the tube. ATSDRAgency for Toxic Substances and Disease Registry — the US public-health agency that publishes toxicological profiles for environmental chemicals's toxicological profile rates oral propylene glycol exposure as low concern at consumer doses; the chemical clears the body within 24 hours via lactate and pyruvate.

Cocamide DEA is the one with the cancer-bioassay record. NTP Technical Report 479 — a two-year dermal study in F344/N rats and B6C3F1 mice — found clear evidence of carcinogenic activity in male and female mice (liver and kidney tumours), equivocal evidencean NTP grade meaning some but not strong evidence — typically a small or marginal increase in tumour incidence in female rats, and no evidence in male rats. The effect was attributed to free DEA contamination of the cocamide condensate NTP TR-479 cocamide DEA bioassay. California Prop 65 listed cocamide DEA as a known carcinogen on 22 June 2012. Most modern toothpaste formulations don't include it — but it's worth a glance at the ingredient list for any term ending in "amide DEA" or "lauramide DEA" if you want to be sure. Uncommon in 2026 toothpaste; not impossible.

Two sweeteners share toothpaste duty, and they've moved in opposite directions. Saccharin spent two decades on the IARC carcinogen list and the NTP Report on Carcinogens before both organisations delisted it — IARC reclassified to Group 3 in 1999 (Volume 73), NTP in the 9th Report on Carcinogens, May 2000 — once the rat-bladder mechanism was shown to be a urine-composition pathway specific to rats and not relevant to humans. Aspartame moved the other way: on 14 July 2023, IARC Volume 134 classified it as Group 2BIARC's classification for agents with limited evidence of carcinogenicity in humans, limited or sufficient in animals — a 'possibly carcinogenic to humans' category that includes aloe vera and pickled vegetables (possibly carcinogenic to humans), citing limited evidence for hepatocellular carcinoma and limited evidence in animals. Either way, dose is the toothpaste-relevant question: a brushing's worth contains a fraction of a milligram, mostly spat out. The systemic dose from toothpaste isn't where the IARC headline applies. If you want to avoid them anyway, fluoride toothpastes for adults are available in unsweetened formulations and most natural lines use stevia, xylitol or nothing at all.

Carrageenan is what makes natural-line toothpastes feel different in the tube. Tom's of Maine fluoride formulations (Whole Care, Cavity Protection, Clean & Gentle) use it; mainstream Colgate, Crest and Sensodyne use carbomer, hydroxyethylcellulose or xanthan gum instead. Tobacman's 2001 review in Environmental Health Perspectives is the canonical concern paper — the argument that food-grade carrageenan can degrade into poligeenan, a recognised animal-model carcinogen, in the acidic stomach environment Tobacman 2001. The food-industry response has been that the degradation doesn't happen at digestive pH and that decades of food use show no human harm signal. The toothpaste exposure is small — most of the paste is spat out, the carrageenan that remains is in oral contact for under a minute — and there's no human study of toothpaste-route carrageenan exposure to draw on either way. Read the back of every tube. "Natural" isn't a defined regulatory category and the formulations vary widely.

Three things to check on the back panel, in roughly the order they matter for the average reader:

Hydroxyapatitethe mineral phase of tooth enamel — used in some non-fluoride toothpastes as a remineralising agent toothpastes are the main fluoride-free alternative with any meaningful evidence base — clinical trials in Japan dating to the 1980s and a smaller European-trial body in the 2010s and 2020s. The honest comparison: hydroxyapatite has a smaller and shorter-duration evidence base than fluoride at 1000–1500 ppm. It's a reasonable swap if the systemic-fluoride concern is decisive for you (pregnancy, early childhood, or you live in a fluoridated water area and want to reduce one exposure source). It isn't yet a like-for-like clinical equivalent. The oral-care chemicals guide covers the hydroxyapatite trial body in more detail.

The eso-friendly call on a toothpaste tube is shorter than the chemistry. Switch for titanium dioxide — cosmetic pigment, no upside to weigh against the EFSA genotoxicity finding. Watch for SLS if you've ever had recurrent mouth ulcers — drop it before reaching for medication. OK for fluoride at 1000–1500 ppm in non-pregnant adults — the dental-evidence base is one of the strongest in the consumer-health literature. The rest of the back panel is formulation chemistry that's fine in trace doses and unnecessary if a simpler tube does the same job.

Twist the cap, read the back, decide. The eight ingredients above are most of what's interesting on a toothpaste tube — and the regulatory inconsistency they sit on top of is the deeper story. Banned from food, legal in mouths. Banned from soap, legal between teeth. Same regulator, different aisle, different rules. For the broader routine — mouthwash, floss, toothbrushes — continue with the full oral-care picture. For the contested-fluoride case in depth, the fluoride profile is the next stop.