Ingredients
How fluoride works
Fluoride works, and the evidence is strong. It works by sitting on the surface of a tooth and helping it rebuild, not by being swallowed. We make a fluoride-free toothpaste and this page still says so.
It works on the outside of the tooth
For most of the last century the explanation was that children needed to swallow fluoride so it could be built into teeth that were still forming. That turned out to be wrong: the amount reaching a developing tooth that way is far too small to explain the benefit. Researchers settled it by the late 1990s. Almost all of the benefit comes from fluoride touching a tooth that has already come in.
What it actually does
Teeth constantly lose mineral to acid and get it back from saliva. Fluoride tilts that trade two ways. Mineral that rebuilds with fluoride present forms a tougher surface. And fluoride leaves a thin deposit that works like a battery, releasing fluoride when acid arrives, which is exactly when it is needed.
How good is the evidence?
Very good. The largest review pooled 74 studies covering more than 42,000 people and found that fluoride toothpaste clearly prevents cavities compared with toothpaste without it. This is not a contested question.
Who actually benefits, and by how much
When a study reports that fluoride cut cavities by some percentage, that is an average across everyone who took part. Almost nobody is average.
Trials enrol broad samples of the public without sorting people by how prone to decay they are, because sorting them is genuinely difficult to do well. So the result blends two very different groups. People at high risk of cavities get a large benefit. People at low risk get a small one, and some of them get little or nothing at all. The average lands in between and describes neither.
A rough way to place yourself: count the cavities you have had, as a child and since. More fillings means higher risk, and higher risk means more to gain. Dry mouth, frequent snacking and gum recession push the same way. Better still, ask your dentist, who can see your history.
Something matters more than fluoride
Fluoride toothpaste cuts new decay by about a quarter. That is the pooled figure from 70 studies covering 42,300 children, and it is a good result for a toothpaste.
Now put that next to what happens when saliva fails. Among people whose salivary glands are damaged by radiation treatment, roughly half develop new cavities within 6 to 12 months, averaging almost four teeth each. That is not a quarter more decay. It is a different disease, arriving in under a year.
Saliva is what actually protects your teeth, as covered in saliva and oral health. Fluoride helps that process along rather than replacing it.
So a person who develops dry mouth is in far more trouble than a person who simply never uses fluoride toothpaste. Nobody has run the head-to-head study, but the sizes involved are not close.
Persistent dry mouth also puts you in the higher-risk group, so it is worth raising with a dentist. What they recommend for it is usually stronger than anything sold over the counter.
What about fluorosis?
Dental fluorosis is faint white flecking, occasionally visible mottling. It depends on dose and only forms while teeth are developing, so it cannot start in an adult mouth. How much toothpaste a child swallows is the main lever, covered in how much toothpaste to use.
Does fluoride irritate your gums?
Rarely, and less often than the internet suggests. Fluoride appears on the list of ingredients that have caused reactions, but near the bottom of it. The usual culprits are flavourings, especially mint and cinnamon, then the foaming agent and preservatives.
Where fluoride has been blamed there is a pattern. The two largest collections of cases, published 67 years apart, both involved stannous fluoride, which contains tin. In one case allergy testing pointed at the tin rather than the fluoride. That does not clear fluoride, but it suggests the tin does much of the work.
So if toothpaste makes your mouth sore, switch to a different flavour before you drop fluoride. If it keeps happening, ask a dentist to test what you are reacting to.
What this page is not saying
- It is not saying fluoride does not work. It does, and the evidence is strong.
- It is not saying fluoride-free is better for your teeth. We make one because some people want one.
- It does not cover fluoride in drinking water, which has its own efficacy and safety pages.
- Whether fluoride is right for you depends on your risk, age and history. Ask a dentist.
Read the full evidence review The mechanism in full, with 9 cited sources and a plain account of where the evidence is weak.
We make a fluoride-free toothpaste, so it is worth stating plainly at the top: fluoride toothpaste works, the evidence that it works is among the strongest in dentistry, and this article is not an argument against it. It is an explanation of the mechanism, which is widely misunderstood, including by people who sell alternatives. Most of what fluoride does happens at the surface of an erupted tooth, not inside a developing one.
Key points
- Fluoride's benefit is overwhelmingly topical. The older model, in which swallowed fluoride built stronger teeth during development, was largely abandoned by the end of the 1990s.
- Fluoride works mainly by shifting the demineralization and remineralization balance at the tooth surface, favoring redeposition of mineral in a less soluble form.
- It forms calcium fluoride-like deposits on enamel that act as a reservoir, releasing fluoride when pH falls, which is exactly when it is needed.
- It also inhibits bacterial glycolysis and acid tolerance, though this is generally considered a secondary contribution at toothpaste concentrations.
- Cochrane reviews find a clear caries-preventive effect for fluoride toothpaste against non-fluoride toothpaste, with a dose-response relationship above 1000 ppm. Dental fluorosis is a real, dose-dependent effect of exposure during tooth development.
The model that was abandoned
For decades fluoride was explained systemically. The story was that fluoride ingested during childhood was incorporated into the crystal structure of developing enamel, producing a tooth that was intrinsically more acid resistant when it erupted. On that model, the point of fluoride was to swallow it while teeth were forming.
That explanation did not survive scrutiny. The concentration of fluoride achievable in enamel through systemic exposure turns out to be far too low to account for the observed reduction in caries, and the benefit tracks ongoing topical exposure rather than developmental history. Reviews through the late 1990s converged on the conclusion that the dominant mechanisms are post-eruptive and topical.1,2
This is not a fringe revision. It is the mainstream position, and it has a practical consequence worth stating: on the current understanding, fluoride does its work while it is in contact with the tooth, which is an argument about where fluoride is delivered rather than about whether it is effective.
Mechanism one: shifting the mineral balance
As covered in how cavities form, enamel dissolves when the fluid against it becomes undersaturated with respect to tooth mineral, and rebuilds when it is supersaturated. Fluoride intervenes on both sides of that exchange.
When mineral is redeposited in the presence of fluoride, fluoride ions substitute for hydroxyl ions in the apatite lattice, producing fluorhydroxyapatite. In practice the product is a mixed phase rather than pure fluorapatite, and the important property is that it is less soluble in acid than the hydroxyapatite it replaces. A surface that has remineralized in the presence of fluoride therefore withstands a lower pH before it begins dissolving again.3
Fluoride also adsorbs to the crystal surface and interferes with dissolution directly, slowing mineral loss during an acid episode. The net effect across a Stephan curve is that less mineral leaves on the way down and more returns on the way up.
A caution on a number you will see quoted. It is often claimed that fluoride lowers the critical pH from 5.5 to about 4.5. The direction is right and the mechanism is real, but as explained in our cavities article, critical pH is not a fixed constant. It depends on the calcium and phosphate content of the fluid at the tooth surface, so treat these figures as illustrative rather than exact.
Mechanism two: the reservoir
When a fluoride product contacts enamel at sufficient concentration, it deposits a layer of calcium fluoride-like material on the surface. This is not part of the enamel crystal. It is a loosely bound deposit, stabilized by phosphate and protein from saliva, sitting on top of the tooth and within plaque.4
What makes it useful is its pH sensitivity. The deposit is relatively stable at neutral pH and dissolves as pH falls, releasing free fluoride into the immediate environment precisely during an acid challenge. It behaves, in effect, as a slow-release store that discharges on demand and is recharged by the next exposure.
This reservoir mechanism is a large part of why frequency of fluoride exposure matters more than concentration at any single moment, and why the practical advice attached to fluoride toothpaste concerns regular contact rather than intensity.
Mechanism three: effects on bacteria
Fluoride has genuine antibacterial activity, and the biochemistry is well characterized. Hydrogen fluoride is uncharged and crosses bacterial membranes freely; inside the more alkaline cytoplasm it dissociates, acidifying the cell interior and releasing fluoride ions where they can act.5
Two targets are usually named. Fluoride inhibits enolase, an enzyme in the glycolytic pathway, which reduces the bacterium's ability to metabolize sugar into acid. It also inhibits the proton-pumping ATPase that acid-tolerant organisms use to expel protons and survive a low-pH environment. Together these reduce both acid production and acid tolerance.6
The honest qualification is that much of this work was done at concentrations and exposure times that do not straightforwardly represent two minutes of brushing, and the antibacterial contribution is generally regarded as secondary to the physicochemical mechanisms above. Presenting fluoride primarily as an antibacterial agent overstates the case.
One important exception sits outside this discussion. Stannous fluoride has a substantial and well-evidenced antibacterial effect on gingival inflammation, but the agent responsible is the tin rather than the fluoride, and the indication is gum health rather than caries. That is covered separately in stannous fluoride and gum health.
What the trials show
The Cochrane review of fluoride toothpaste in children and adolescents, covering seventy-four studies and more than forty-two thousand participants, found a clear caries-preventive benefit against non-fluoride toothpaste.7 This is one of the better evidenced interventions in preventive dentistry, and it is not seriously contested.
A separate Cochrane review examined concentration. It found that toothpastes at 1000 ppm and above outperform non-fluoride toothpaste, with evidence of a dose-response relationship, while the evidence for lower concentrations was weaker and less certain.8 Most standard adult toothpaste sits between 1000 and 1500 ppm.
Worth noting for anyone reading effect sizes: these trials are typically reported as a prevented fraction in the increment of decayed, missing and filled surfaces over the study period. That is a relative measure, and the absolute benefit to a given person depends heavily on their underlying caries risk.
That last point deserves more weight than it usually gets, because it is the difference between a population result and a personal one. Caries risk is not distributed evenly. It clusters, and the distribution is heavily skewed: a minority of individuals carry a disproportionate share of disease. Trials of dentifrice recruit broad samples and do not stratify by caries risk, partly because reliable individual-level risk prediction remains poor, as noted in how cavities form. The pooled figure is therefore not an estimate of the benefit to any particular person, and no published trial reports what it would be.
A pooled prevented fraction is therefore an average over a mixed population containing people with a great deal to gain and people with very little. Applied to a low-risk individual it overstates the expected benefit; applied to a high-risk one it understates it, potentially by a wide margin. Neither error is visible in the headline figure, and the same caution applies to every dentifrice active, including the fluoride-free ones.
The practical version. Past caries experience remains the strongest available predictor of future caries, which makes a personal filling count a crude but genuinely informative proxy: the more restorations you have accumulated, the more a caries-preventive agent is likely to be doing for you. Reduced salivary flow, frequent fermentable carbohydrate exposure and exposed root surfaces all shift the same way. Formal caries risk assessment performed by a dentist is the right instrument, and it is the reason fluoride recommendations are increasingly individualized rather than uniform.
Salivary function dominates the caries balance
A page about fluoride should say where fluoride sits in the hierarchy of things that determine whether enamel survives, and it does not sit at the top.
The pooled prevented fraction for fluoride dentifrice against a non-fluoride control is 24 percent (95% CI 21 to 28), from 70 studies and 42,300 children.16 That is a reduction of roughly a quarter in the caries increment, and it is among the better-evidenced effects in preventive dentistry.
Set that against the loss of salivary function. In head and neck cancer patients following radiotherapy, approximately 49 percent develop new carious lesions within 6 to 12 months, averaging 3.7 teeth affected.17 Against a typical background increment of around 2.6 D(M)FS per year in the populations Cochrane pooled, that is a change in kind rather than degree. In primary Sjögren's syndrome, meta-analysis of cross-sectional studies reports a mean DMFS difference of 4.42 (95% CI 2.44 to 6.41) against controls, with caries appearing at atypical sites including incisal and root surfaces.18
The mechanism is covered in saliva and oral health and is not in dispute. Losing salivary flow removes the mineral reservoir, the bicarbonate buffer, the clearance mechanism, the pellicle supply and the antimicrobial protein load simultaneously. Fluoride modifies one term in the caries balance described in how cavities form. Saliva supplies several of the others at once.
Stated plainly, and with the caveat it needs: an individual who develops significant hyposalivation is at substantially greater caries risk than an individual with normal salivary function who uses no fluoride toothpaste at all. No study has compared those two groups directly and none is likely to, so this is an inference from effect magnitudes rather than a measured contrast. The magnitudes are not close, and severe hyposalivation is described in the clinical literature as one of the strongest caries risk factors there is, operating independently of how carefully someone brushes.
The conclusion that follows is the opposite of the one people expect. Because caries risk is skewed, and because fluoride's absolute benefit scales with underlying risk, the hyposalivic patient is precisely the person for whom fluoride does the most work. Clinical practice reflects this: high-concentration fluoride dentifrice is a standard part of managing radiation-induced and drug-induced dry mouth. Salivary function dominating the hierarchy and fluoride mattering enormously to dry-mouth patients are the same fact viewed from two directions, not competing claims.
The practical implication is about priority. Anyone weighing which toothpaste to buy while also taking a medication that dries their mouth, mouth-breathing at night, or noticing persistent dryness has the ordering wrong. The tube is a smaller lever than the underlying salivary problem, and the latter is a clinical question rather than a shopping one.
Dental fluorosis
Dental fluorosis is a developmental effect, not a reaction in an adult mouth. It arises when fluoride exposure during the years of enamel formation disturbs the maturation of the enamel matrix, producing enamel with altered mineral content. Clinically it ranges from faint white flecking, which is common and usually noticed only by a clinician, through to visible mottling and, at the severe end, pitting.
A 2024 Cochrane review examined topical fluoride as a cause of fluorosis in children. It found the evidence limited and generally of low certainty, with the clearest signal for fluorosis of aesthetic concern associated with starting fluoride toothpaste before roughly twelve months of age.9 The practical guidance that follows from this concerns supervising the amount of toothpaste young children use and discouraging swallowing, not avoidance.
Fluorosis is the reason fluoride recommendations are age-stratified. It is a dose-dependent effect with a defined window of susceptibility, and it belongs in an honest account of the mechanism.
Soft tissue reactions, and why the signal points at tin
Toothpaste causes contact reactions in the mouth. That is established, and covered in our articles on surfactants and the oral mucosa and preservatives and the oral mucosa. The question here is narrower: how much of it is attributable to fluoride.
Clinical dermatology references are consistent on the ranking. Flavourings are the major cause, with mint-derived compounds including spearmint, peppermint, menthol and carvone named as the most frequent responsible substances, followed by cinnamal. Roughly 30 potential allergens have been identified in dentifrice, and fluoride salts appear on that list without any accompanying detail on frequency or potency.10 Reactions overall are described as rare relative to how universally these products are used.
The fluoride-specific literature is thin, old and interesting. In 1957 Thomas Douglas, a Seattle physician, reported 133 patients who developed stomatitis after using fluoride dentifrice, and in 32 of them reproduced the lesions by reintroducing the paste, in some cases as many as six times.11 Withdrawal and rechallenge is a genuinely meaningful design element, and it is the strongest feature of an otherwise weak report: uncontrolled, unblinded, single-observer, published in a regional journal, and never replicated at that scale.
In 2024 a French group reported 15 patients with eruptive white oral lesions, all users of toothpaste containing stannous fluoride, describing it as the first case series of its kind.12 There is also a published case of allergic contact cheilitis in which patch testing identified tin, the metal in stannous fluoride, as the responsible allergen rather than the fluoride ion.13
A detail that connects these and is worth stating explicitly. Crest launched in 1955 as a stannous fluoride formulation, branded Fluoristan, and Procter & Gamble did not switch it to sodium fluoride until 1981.19 Douglas's 1957 patients were therefore using stannous fluoride toothpaste, as were the 2024 French patients. The two largest case collections in this literature, separated by 67 years, implicate the same salt, and the one case with a positive patch test identifies the tin rather than the fluoride. That is a pattern consistent with the tin being responsible. It is not proof, since nobody has run the controlled comparison, but it is the most parsimonious reading of what exists.
Two further pieces of evidence complicate any simple story. Patch testing with sodium fluoride and stannous fluoride at dentifrice concentrations is reported to be pro-inflammatory on previously damaged skin while producing no inflammation on intact skin, which suggests an irritant rather than allergic mechanism and makes barrier status the relevant variable. And a 2013 case report attributes perioral dermatitis to a 5000 ppm prescription-strength dentifrice, with resolution on discontinuation14, at a concentration three to five times that of an over-the-counter product.
The evidence also runs firmly in the opposite direction on gingival health specifically. Stannous fluoride is used as an anti-gingivitis active because it works: randomized trials report significantly reduced gingival bleeding and inflammation against control dentifrices, attributed to reduced biofilm viability.15 Both things can be true at once, and probably are. A population-level reduction in gingival inflammation is entirely compatible with idiosyncratic reactions in a small minority, and conflating the two produces a false picture in either direction.
The honest summary: fluoride can cause oral soft tissue reactions, the evidence for it is case reports rather than controlled data, it is well down the list of dentifrice ingredients that do so, and the signal that exists tracks the stannous salt more closely than the fluoride ion. Anyone with recurrent oral soreness after brushing has a reasonable case for changing product and a better case for asking a clinician to patch test, and very little basis for concluding that fluoride specifically is the cause.
What this page does not argue
- It does not argue that fluoride is ineffective. The evidence that fluoride toothpaste reduces caries is strong, consistent and long-standing. Any page suggesting otherwise is misrepresenting the literature.
- It does not argue that fluoride-free is superior. We make a fluoride-free toothpaste because some people want one. That is a preference we serve, not a claim of better clinical outcomes, and we are not aware of evidence that would support such a claim.
- It does not cover water fluoridation. How much water fluoridation still contributes once toothpaste is in use is treated in fluoridated water versus toothpaste. Questions about water fluoridation, total exposure and neurodevelopmental research are a separate subject with a separate and more contested evidence base. Mixing them into a mechanism article would do justice to neither.
- It is not advice about your own teeth. Whether fluoride is appropriate for you depends on your caries risk, your age, your salivary function and your history. That is a conversation with a dentist.
References
- ten Cate JM. Current concepts on the theories of the mechanism of action of fluoride. Acta Odontologica Scandinavica. 1999;57(6):325-329. PMID 10777135.
- ten Cate JM, Buzalaf MAR. Fluoride mode of action: once there was an observant dentist. Journal of Dental Research. 2019;98(7):725-730. PMID 31219410.
- Buzalaf MAR, Pessan JP, Honório HM, ten Cate JM. Mechanisms of action of fluoride for caries control. Monographs in Oral Science. 2011;22:97-114. PMID 21701194.
- Øgaard B, Seppä L, Rølla G. Professional topical fluoride applications: clinical efficacy and mechanism of action. Advances in Dental Research. 1994;8(2):190-201. PMID 7865075.
- Marquis RE, Clock SA, Mota-Meira M. Fluoride and organic weak acids as modulators of microbial physiology. FEMS Microbiology Reviews. 2003;26(5):493-510. PMID 12586392.
- Marquis RE. Antimicrobial actions of fluoride for oral bacteria. Canadian Journal of Microbiology. 1995;41(11):955-964. PMID 7497353.
- Marinho VCC, Higgins JPT, Sheiham A, Logan S. Fluoride toothpastes for preventing dental caries in children and adolescents. Cochrane Database of Systematic Reviews. 2003;(1):CD002278. PMID 12535435.
- Walsh T, Worthington HV, Glenny AM, Marinho VCC, Jeroncic A. Fluoride toothpastes of different concentrations for preventing dental caries. Cochrane Database of Systematic Reviews. 2019;3(3):CD007868. PMID 30829399.
- Wong MCM, Zhang R, Luo BW, et al. Topical fluoride as a cause of dental fluorosis in children. Cochrane Database of Systematic Reviews. 2024;6(6):CD007693. PMID 38899538.
- Contact reactions to toothpaste and other oral hygiene products. DermNet. DermNet.
- Douglas TE. Fluoride dentifrice and stomatitis. Northwest Medicine. 1957;56:1037-1039.
- Vinant M, et al. Oral contact stomatitis related to toothpaste use: a report of 15 cases. JEADV Clinical Practice. 2024. doi:10.1002/jvc2.318.
- Cheilitis and urticaria associated with stannous fluoride in toothpaste, with tin identified on patch testing. Dental Update. Dental Update.
- Peters P, Drummond C. Perioral dermatitis from high fluoride dentifrice: a case report and review of literature. Australian Dental Journal. 2013;58(3):371-372. PMID 23981221.
- Stannous Fluoride in Toothpastes: A Review of Its Clinical Effects and Likely Mechanisms of Action. 2025. PMC11942899. See also the 3-week randomised gingival health study, PMC8436562.
- Marinho VCC, Higgins JPT, Sheiham A, Logan S. Fluoride toothpastes for preventing dental caries in children and adolescents. Cochrane Database of Systematic Reviews. 2003;(1):CD002278. Pooled D(M)FS prevented fraction 24% (95% CI 21 to 28), 70 studies, 42,300 children. Cochrane Library.
- Dental caries following radiotherapy for head and neck cancer: a systematic review. Oral Oncology. 2019. Oral Oncology. See also Over 300 Radiation Caries Papers: Reflections From the Rearview Mirror, Frontiers in Oral Health. 2022;3:961594, PMC9330023.
- Salivary changes and dental caries as potential oral markers of autoimmune salivary gland dysfunction in primary Sjögren's syndrome; and meta-analysis of DMFS in Sjögren's disease versus controls. PMC554998.
- American Chemical Society. National Historic Chemical Landmark: the development of Crest toothpaste, marketed from 1955 with stannous fluoride under the Fluoristan trademark and reformulated to sodium fluoride, as Fluoristat, in 1981. ACS. See also the Smithsonian National Museum of American History object record for Crest with Fluoristat, NMAH 688610.
This article is educational and describes published research on the mechanism of action of fluoride. It is not medical or dental advice, not a diagnosis, and not a claim about any product. Decisions about fluoride use should be made with a dentist who knows your history.
Keep reading
Athena Naturals
Seven ingredients. Fluoride-free, no water, no preservatives, no surfactants. Made in small batches in Seattle.