Evidence review
Fluoride in drinking water
Fluoride is a dose problem, not a yes-or-no problem. At the level American water is fluoridated, the current evidence does not show a risk to IQ or to teeth. The concerns that do exist are real, and they begin at higher doses.
Three numbers worth knowing
0.7 mg/L is what US health authorities recommend for fluoridated tap water. 1.5 mg/L is the World Health Organization's guideline ceiling. 4.0 mg/L is the legal maximum the EPA allows, and it was set to prevent crippling bone disease, not to be a target.
These do different jobs. One is a dose aimed at reducing cavities. One is a safety guideline. One is the point at which water is legally unsafe. They get quoted interchangeably, and that is where most of the confusion starts.
What the IQ research found
In 2024 the US National Toxicology Program reviewed the evidence and concluded, with moderate confidence, that fluoride exposure above 1.5 mg/L is associated with lower IQ in children. Of 19 studies it rated high quality, 18 pointed the same way. That is a serious finding and it should not be brushed aside.
Two things about it matter. The studies came mostly from places where people were drinking more fluoride than Americans do, including parts of China, India, Iran, Pakistan, Mexico and Canada. And the review said plainly that it could not determine whether there is an effect below 1.5 mg/L. US water is fluoridated at less than half that concentration.
The court case people cite
In September 2024 a federal court in California ruled that fluoridating water at 0.7 mg/L poses an unreasonable risk to children's IQ, and ordered the EPA to act.
That ruling no longer stands. On 21 May 2026 the Ninth Circuit vacated it, on the grounds that the trial judge had improperly taken over how the case was argued. The appeal court did not rule on whether fluoride is safe. So there is currently no federal finding either way, and a lot of pages online still describe the 2024 decision as though it were settled law.
Fluorosis, in teeth and in bone
Dental fluorosis is the effect that does happen at ordinary exposures. It only forms while teeth are still developing in childhood, so it cannot start in an adult mouth. Most of it is faint white flecking that only a dentist notices. At the severe end it becomes visible mottling and pitting.
It is not only cosmetic. Fluorotic enamel is measurably weaker: roughly 10 percent lower surface hardness, with a porous, under-mineralized layer beneath. More fluoride does not build a stronger tooth past a point. It builds a more porous one.
Skeletal fluorosis, which stiffens and embrittles bone, needs something in the range of 10 to 20 mg of fluoride a day for 10 to 20 years. That is why the EPA's legal ceiling sits at 4.0 mg/L. It is not reachable from water at 0.7 mg/L, and it is not reachable from toothpaste.
How to look up your own water
This is public information and you are entitled to it. Every water utility in the country has to publish an annual Consumer Confidence Report listing what is in the water, fluoride included. The CDC also runs a lookup tool called My Water's Fluoride covering roughly 40 states.
Two caveats. Neither source is live, so treat the figure as a recent average rather than today's number. And if you are on a private well, nobody is testing it for you. Well water is where genuinely high natural fluoride concentrations turn up in the US, and the only way to know is to have it tested.
Where this leaves it
- At 0.7 mg/L in water, and with over-the-counter toothpaste used as directed, current research does not support a risk to IQ or a meaningful risk of fluorosis.
- The IQ findings are real, but they sit at more than twice the US water concentration, and the effect below 1.5 mg/L is genuinely unknown rather than ruled out.
- Dental fluorosis is the effect that does occur at US exposures. It comes from total intake during childhood, which is why how much toothpaste a child swallows matters.
- Look up your own water. It takes two minutes and the information is already published.
Read the full evidence review The regulatory numbers, the neurodevelopmental literature, the vacated court ruling, and where the evidence runs out.
We make a fluoride-free toothpaste, so the bias here runs one way and readers should weigh it accordingly. This page is not an argument that fluoride is dangerous. It is an attempt to put the actual numbers next to each other, because almost every public argument about fluoride is really an argument about dose being conducted without one. The mechanism by which fluoride works is covered separately in how fluoride works.
Key points
- US community water is fluoridated at 0.7 mg/L. The WHO guideline value is 1.5 mg/L and the EPA's enforceable maximum is 4.0 mg/L, a limit set to prevent crippling skeletal fluorosis rather than to define an optimum.
- The 2024 National Toxicology Program monograph found, with moderate confidence, an association between fluoride above 1.5 mg/L and lower IQ in children. It explicitly could not determine whether an effect exists below that concentration.
- The 2024 district court ruling that fluoridation poses an unreasonable risk was vacated by the Ninth Circuit on 21 May 2026 on procedural grounds. It is not current law and the appeal did not address the science.
- Dental fluorosis is developmental and dose-dependent. Fluorotic enamel is measurably weaker, with roughly 10 percent lower nanohardness and a porous subsurface.
- Skeletal fluorosis requires roughly 10 to 20 mg per day for 10 to 20 years. It is not reachable from fluoridated water or from toothpaste used as directed.
Four numbers that are constantly confused
Most fluoride arguments collapse the moment the concentrations are written down side by side, because the figures being traded are answers to different questions.
| Figure | Value | What it is |
|---|---|---|
| US Public Health Service recommendation | 0.7 mg/L | Target concentration for community water fluoridation, set in 2015, replacing the 1962 range of 0.7 to 1.21 |
| WHO guideline value | 1.5 mg/L | Health-based guideline adopted as the national standard in most countries2 |
| EPA maximum contaminant level | 4.0 mg/L | Enforceable US limit, set to protect against crippling skeletal fluorosis3 |
| IOM tolerable upper intake, adults | 10 mg/day | Total intake from all sources; for children through age 8 the limit is 0.10 mg per kg of body weight per day4 |
The distinction between the third and fourth rows is the one that does the most work. A concentration limit says nothing on its own about a given person's dose, which depends on how much they drink, their body weight, and what else they are swallowing. An adult drinking two liters of water at 0.7 mg/L takes in about 1.4 mg per day, roughly 14 percent of the tolerable upper intake.
Coverage is also narrower than most people assume. In 2022, 62.8 percent of the US population received fluoridated water, or 72.3 percent of those on community water systems.5 That figure has since fallen: Utah and Florida both enacted statewide bans in May 2025.6
The neurodevelopmental evidence
The 2024 National Toxicology Program monograph is the most rigorous synthesis available and is the document worth reading rather than any coverage of it. Its conclusion, stated precisely: moderate confidence that higher fluoride exposure, such as drinking water above 1.5 mg/L, is associated with lower IQ in children. Of 19 studies rated high quality, 18 reported an inverse association.7
Three qualifications belong with that sentence, and omitting any of them misrepresents the document.
First, the underlying studies were conducted largely outside the US, in Canada, China, India, Iran, Pakistan and Mexico, in populations where total fluoride exposure exceeded 1.5 mg/L. Second, the monograph states that more studies are needed to understand whether exposure below 1.5 mg/L affects IQ. That is a statement of uncertainty, not of safety, and it should not be reported as either. Third, "moderate confidence" is a defined term in systematic review methodology, and it sits below "high confidence".
The honest reading is narrow and worth stating exactly. There is reasonable evidence of harm above 1.5 mg/L. There is an absence of adequate evidence below it. An absence of evidence at 0.7 mg/L is not the same as evidence of absence, and anyone presenting it as either a clean bill of health or a demonstrated harm has gone past what the document supports.
Food & Water Watch v. EPA, and what happened to it
In September 2024 the US District Court for the Northern District of California found that fluoridation at 0.7 mg/L presents an unreasonable risk of reduced IQ in children under the Toxic Substances Control Act, and directed the EPA to take regulatory action.8 It was the first citizen petition under TSCA section 21 to reach trial and win, and it was reported worldwide.
On 21 May 2026 the Ninth Circuit vacated that decision and remanded the case.9 The vacatur rested on procedure rather than science: the appeal court found the district judge had improperly interfered in how the parties presented their case, admitting and relying on evidence the parties had agreed should not be considered, and insisting on a second trial that neither side had requested.
The practical position, as of this review date, is that no federal court finding on fluoride's neurotoxicity stands. The appellate decision is not an endorsement of fluoridation's safety either. A large share of material published online still cites the 2024 ruling as current, which is worth knowing when reading anything on this subject.
Dental fluorosis
Fluorosis arises when fluoride exposure during enamel formation disturbs maturation of the enamel matrix. It is developmental, dose-dependent, and confined to a window of susceptibility in childhood. It cannot begin in an adult mouth.
Prevalence in the US has risen substantially. Across three national surveys, total fluorosis prevalence in adolescents aged 12 to 15 went from 22 percent in 1986 to 1987, to 41 percent in 1999 to 2004, to 65 percent in 2011 to 2012.10 Combined moderate and severe fluorosis rose from 1.2 percent to 3.7 percent to 30.4 percent across the same surveys. Treat that last progression with some caution, since examiner training and diagnostic criteria were not identical across surveys separated by 25 years, and a share of the increase is likely ascertainment. The direction of travel is nonetheless not in dispute.
The consequence is not purely cosmetic, which is how it is usually described. Fluorotic enamel is structurally different: excess retained matrix protein, hypomineralized enamel rods, and widened spaces between them producing subsurface porosity. Nanoindentation puts the nanohardness of fluorosed enamel at roughly 10 percent below normal enamel.11
This is the point where the popular claim that fluoride "hardens" teeth stops being useful. Fluoride incorporated during remineralization does produce a less acid-soluble mineral, which is the mechanism described in how fluoride works. Excess fluoride during development produces the opposite: a softer, more porous enamel. Same element, opposite structural outcome, and the variable separating them is dose and timing.
One claim we could not support is worth naming, because we went looking for it. There is no good evidence that fluoride toothpaste at normal concentrations makes enamel more brittle. Fluoride varnish raises surface microhardness by depositing calcium fluoride, and sodium fluoride toothpaste has been measured to produce no obvious change in crack length, fracture toughness or brittleness. Increased brittleness is a property of fluorotic enamel from excess ingestion during development, not of topical fluoride use.
Skeletal fluorosis and fracture risk
Fluoride is bone-seeking. Chronic excess produces skeletal fluorosis: increased bone density accompanied by reduced bone quality, calcification of ligaments, joint stiffness and, at the crippling end, skeletal deformity and neurological deficit from spinal cord compression.
The dose required is large and sustained. The National Research Council put crippling skeletal fluorosis at 10 to 20 mg per day for 10 to 20 years.12 The IOM's 10 mg/day upper intake level derives from the same body of evidence. Endemic skeletal fluorosis occurs where groundwater carries very high natural fluoride, with Indian case series describing onset after around 10 years of residence at roughly 9 mg/L.
Fracture risk at lower exposures is a separate and more contested question. Dose-response meta-analysis reports a non-linear relationship with an approximate threshold around 1.5 mg/L, above which fracture risk rises roughly linearly.13 Below that threshold the picture is not clear, and this literature is observational.
For a US reader on fluoridated water, neither endpoint is within reach at 0.7 mg/L. Both are relevant to well water in high-fluoride geology, which is the exposure route that actually produces these conditions in this country.
Finding your own water's concentration
This is published information and takes very little effort to retrieve.
- Your Consumer Confidence Report. Every community water system in the US is required to publish one annually, listing detected contaminants and fluoride concentration. Your utility's website will have it.
- CDC My Water's Fluoride. A lookup tool covering roughly 40 participating states.14 The CDC notes it does not carry real-time data, so read it as a recent typical value.
- Your utility directly. The CDC's own guidance is that the local water provider is the best source.
Private wells are outside all of this. No agency monitors them, no report is issued, and naturally occurring fluoride in US groundwater varies widely, running higher west of the Mississippi and in parts of the Southwest and eastern Midwest. If you are on a well and the question matters to you, the only answer is a laboratory test of your own water.
Where the evidence is strong and where it is not
Strong. The dose-dependence of dental fluorosis, and its mechanism. The exposure required for skeletal fluorosis. The structural and mechanical differences between fluorotic and normal enamel. The regulatory figures, which are simply published values.
Moderate. The association between fluoride above 1.5 mg/L and lower childhood IQ, which is the NTP's own characterization of its finding.
Weak or absent. Any conclusion about neurodevelopmental effects at 0.7 mg/L, in either direction. Fracture risk below 1.5 mg/L. The magnitude of the apparent rise in moderate-to-severe fluorosis, given changing survey methodology.
Not addressed here. How much benefit water fluoridation still delivers, which is covered in fluoridated water versus toothpaste. Whether community water fluoridation is good policy. That question involves caries reduction, cost, equity of access to dental care and individual consent, and it is not answerable from toxicology alone. We have not tried to answer it, and a company selling a fluoride-free product is not the right party to.
References
- US Department of Health and Human Services Federal Panel on Community Water Fluoridation. U.S. Public Health Service Recommendation for Fluoride Concentration in Drinking Water for the Prevention of Dental Caries. Public Health Reports. 2015;130(4):318-331. HHS.
- World Health Organization. Guidelines for Drinking-water Quality. Guideline value for fluoride, 1.5 mg/L.
- US Environmental Protection Agency. Fluoride in Drinking Water. Maximum contaminant level 4.0 mg/L. EPA.
- Institute of Medicine. Fluoride. In: Dietary Reference Intakes for Calcium, Phosphorus, Magnesium, Vitamin D, and Fluoride. National Academies Press, 1997. NBK109832.
- Centers for Disease Control and Prevention. Water Fluoridation Data and Statistics, 2022. CDC.
- Utah HB 81 and Florida SB 700, both signed 7 May 2025, ending community water fluoridation statewide.
- National Toxicology Program. NTP Monograph on the State of the Science Concerning Fluoride Exposure and Neurodevelopment and Cognition: A Systematic Review. NIEHS, August 2024. NBK606081; PMID 39172715.
- Food & Water Watch, Inc. v. EPA, No. 17-cv-02162-EMC, 2024 WL 4291497 (N.D. Cal. 24 September 2024).
- Food & Water Watch, Inc. v. EPA, No. 25-384 (9th Cir. 21 May 2026), memorandum disposition vacating and remanding.
- Neurath C, Limeback H, Osmunson B, Connett M, Kanter V, Wells CR. Dental Fluorosis Trends in US Oral Health Surveys: 1986 to 2012. JDR Clinical & Translational Research. 2019;4(4):298-308. PMID 30931722.
- Investigation on the Gradient Nanomechanical Behavior of Dental Fluorosis Enamel. Nanoscale Research Letters. 2018;13:347. PMC6207606.
- National Research Council. Fluoride in Drinking Water: A Scientific Review of EPA's Standards. National Academies Press, 2006. NAP.
- The association of fluoride exposure with bone density and fracture risk: a dose-response meta-analysis. PMID 41068882.
- Centers for Disease Control and Prevention. My Water's Fluoride. CDC MWF.
This article is educational and describes published research and public regulatory standards. It is not medical or dental advice, not a diagnosis, and not a claim about any product. We make a fluoride-free toothpaste and have an interest in this subject, which readers should weigh. Decisions about fluoride for you or your children belong with a dentist or physician who knows your history.
Keep reading
Athena Naturals
Seven ingredients. Fluoride-free, no water, no preservatives, no surfactants. Made in small batches in Seattle.