Water Fluoridation: What the Evidence Now Shows

Oct 8, 2026 · @Thomas Abshier

Community water fluoridation has been treated as settled public health practice in America for eighty years. That is no longer the case. Since 2024, a federal toxicology review, a federal trial, several new meta-analyses, state legislatures, and now the EPA itself have reopened the question of whether adding fluoride to public water is safe for children’s developing brains, and whether it still does enough for teeth to justify adding it at all.

This article lays out what the evidence shows, where it is strong, where it is weak, and what a family can reasonably do while regulators work. It also corrects some claims in the alternative press that go beyond the evidence. A strong case does not need to be overstated, and overstating it gives the other side an easy rebuttal.

How fluoridation began

Grand Rapids, Michigan, became the first city to add fluoride to its water in January 1945. The idea grew from early 1900s observations: dentists in Colorado Springs noticed that residents with brown-mottled teeth, caused by naturally high fluoride in the local water, also had fewer cavities. In the 1930s, H. Trendley Dean of the U.S. Public Health Service compared cities with different natural fluoride levels and concluded that about 1 part per million (1 mg/L) reduced decay without much visible mottling.

From there, fluoridation spread across the country. For decades, the federal recommendation ranged from 0.7 to 1.2 mg/L, depending on climate. In 2015, the Public Health Service lowered it to a single value of 0.7 mg/L, in part because dental fluorosis had become common in American children. Separately, the EPA sets an enforceable maximum of 4.0 mg/L in drinking water and a non-enforceable secondary standard of 2.0 mg/L.

The key fact to hold onto is the date. The original evidence for fluoridation was gathered before fluoride toothpaste existed. Whatever fluoridated water did for teeth in 1945, it did for children who had no other regular fluoride source.

Topical versus systemic: how fluoride actually protects teeth

Fluoride protects teeth mainly by contact with the tooth surface, not by being swallowed. That is now the position of the CDC itself, which stated in 1999 that fluoride’s caries-preventing effect is predominantly topical and occurs after the teeth have erupted.

This matters a great deal. When fluoridation began, the theory was that swallowed fluoride was built into developing enamel and made it more resistant from the inside. If that were the main mechanism, putting fluoride in water would be the logical delivery route. But if the benefit comes from fluoride bathing the tooth surface, where it slows demineralization and helps enamel remineralize, then toothpaste applied directly to the teeth delivers the benefit without requiring anyone to swallow it.

Swallowing is where the risks lie. Ingested fluoride is absorbed into the blood, crosses the placenta, and accumulates in bone. A child who drinks fluoridated water receives the systemic exposure whether or not it does his teeth any good. The modern understanding of the mechanism therefore weakens the original rationale for putting fluoride in water while leaving the case for topical fluoride largely intact.

The neurodevelopmental evidence

The strongest current concern is that fluoride exposure during pregnancy and early childhood may lower children’s IQ. Three bodies of evidence bear on it.

The National Toxicology Program review (August 2024). After a review that took roughly eight years and two rounds of National Academies scrutiny, the NTP concluded with moderate confidence that fluoride exposures above 1.5 mg/L in drinking water, the World Health Organization’s guideline limit, are consistently associated with lower IQ in children (NTP). Moderate confidence is the second-highest of the NTP’s four levels. The NTP was explicit that its data were insufficient to say whether the U.S. level of 0.7 mg/L affects IQ, and that more research is needed below 1.5 mg/L.

The NIH meta-analysis (JAMA Pediatrics, January 2025). Kyla Taylor and colleagues pooled studies from many countries. In studies that measured fluoride in drinking water, the association with lower IQ held below 4 mg/L and below 2 mg/L but was not statistically significant below 1.5 mg/L. In studies that measured fluoride in children’s urine, which captures exposure from all sources, the inverse association held even in the lowest category; among the 11 low-risk-of-bias studies, each 1 mg/L rise in urinary fluoride corresponded to a loss of about 1.14 IQ points (HealthDay summary). The paper drew a sharply critical editorial from public health dentist Steven Levy, who argued that most included studies had a high risk of bias.

Pregnancy cohort studies in fluoridated or low-fluoride settings. The studies most relevant to North American water come from mother–child cohorts that measured fluoride in the mother’s urine during pregnancy. The Canadian MIREC cohort (Green et al., JAMA Pediatrics, 2019) and the Mexican ELEMENT cohort (Bashash et al., Environmental Health Perspectives, 2017) both found that higher prenatal fluoride exposure was associated with lower IQ in the children, at exposure levels comparable to those in fluoridated communities. These were among the studies the federal district court relied on in 2024.

Put together, the evidence is strong that fluoride is a developmental neurotoxicant at higher doses, and suggestive but not settled at the doses Americans receive from fluoridated water. The honest summary is not “fluoridation is proven to lower IQ.” It is that a margin of safety which was once assumed has not been demonstrated, and the most sensitive studies, those measuring actual exposure in pregnant women, point in the wrong direction.

The court case and the EPA review

The 2024 ruling has been vacated. In September 2024, after a bench trial in Food & Water Watch v. EPA, a federal district judge in San Francisco found that fluoridation at 0.7 mg/L presents an unreasonable risk of reduced IQ under the Toxic Substances Control Act and ordered the EPA to act. Many articles still cite that ruling as standing law. It is not. On May 21, 2026, the Ninth Circuit vacated it and sent the case back (National Law Review).

The reason matters. The appeals court did not decide whether fluoride at 0.7 mg/L is safe. It held that the trial judge had overstepped procedurally, by effectively taking over the case and holding a second trial, and it returned the question of the plaintiffs’ standing for reconsideration (ELR summary). The EPA, for its part, appealed on legal grounds and did not ask the court to overturn the trial judge’s findings on the science. The plaintiffs have sought more time to ask the Supreme Court to review the case, with a filing deadline of October 19, 2026 (Supreme Court docket).

The EPA’s own review is moving. Separately from the lawsuit, the EPA is reassessing fluoride under the Safe Drinking Water Act. It released a preliminary assessment plan in January 2026 and a final protocol on August 3, 2026 (EPA protocol). The protocol commits the agency to a dose-response analysis of two outcomes in children, neurodevelopment (including IQ) and dental fluorosis, leading to a reference dose: an estimate of the daily exposure below which harm is unlikely (NGWA). A draft assessment will be released for public comment when complete. No date has been set, and the agency says it will not prejudge the outcome.

Two cautions follow. First, the EPA does not decide whether communities fluoridate; the Safe Drinking Water Act forbids it from requiring fluoridation, so its review can lower the permitted maximum but cannot itself end the practice. Second, a protocol is a plan, not a finding. Readers should be wary of any headline that treats the review’s conclusion as already known in either direction.

How much does fluoridation still help teeth?

The benefit is real but far smaller than the figures quoted for decades. The best current summary is the October 2024 Cochrane review, the standard reference for evaluating medical interventions (Cochrane Oral Health).

Cochrane separated old studies from new ones. Most of the evidence behind the familiar claims of 40 to 60 percent fewer cavities came from studies conducted before 1975, before fluoride toothpaste was in common use. Looking only at studies from after 1975, the reviewers found that starting fluoridation may reduce decay in children’s baby teeth by about a quarter of one tooth per child (0.24 fewer decayed, missing, or filled teeth), with a confidence interval that crosses zero and evidence rated as low certainty (review abstract). They found no qualifying studies in adults, and too little evidence to say whether stopping fluoridation increases decay or whether fluoridation narrows the gap between rich and poor children. A large English study of 6.4 million records (the LOTUS study) reached a consistent conclusion: the benefit has shrunk in the toothpaste era (University of Manchester).

The international picture. Cochrane counts about 25 countries that practice community water fluoridation. In Western Europe, artificial water fluoridation is now limited to parts of the United Kingdom, Ireland, and Spain. Several countries that once fluoridated, including Germany, the Netherlands, Sweden, and Finland, have stopped, and tooth decay in Western Europe has fallen over the past half-century at rates similar to those in fluoridated countries.

Two common claims overstate this, however. First, Western Europe has not “banned” fluoridation; no European country prohibits it, and a 2018 survey of EU health ministries found that countries stopped for practical, ethical, or evidentiary reasons rather than because of documented harm (Dublin City University). Second, Europe has not abandoned fluoride. Several countries, notably Switzerland, Germany, and France, supply it through fluoridated table salt, which lets each household choose. The European lesson is not that fluoride is useless. It is that a population can have good dental health without putting fluoride in the public water.

Other concerns

Dental fluorosis. Fluorosis is visible proof that a child absorbed more fluoride than his developing teeth could handle. Mild forms appear as faint white flecks; severe forms cause brown staining and pitting. National survey data from the 2010s indicate that roughly two-thirds of American adolescents show some degree of fluorosis, most of it mild. It was the main reason the federal recommendation was lowered to 0.7 mg/L in 2015, and it is one of the two outcomes the EPA is now formally assessing. Fluorosis is a cosmetic problem in itself, but it is also a biomarker: it shows that systemic exposure in early childhood is common and often exceeds what was intended.

The dose cannot be controlled. A drug is prescribed in a dose suited to the patient. Fluoridated water delivers a fixed concentration, so the dose depends on how much a person drinks. Infants fed formula made with tap water receive far more fluoride per pound of body weight than adults; for that reason the American Dental Association has long advised parents that low-fluoride water may be used for formula to reduce fluorosis risk. Manual laborers, athletes, people in hot climates, and people with kidney disease, who clear fluoride less efficiently, also receive higher doses. Fluoride arrives from toothpaste, tea, processed foods and beverages made with fluoridated water, and some pesticide residues as well, so the water is only one source.

What is actually added. Most U.S. systems use fluorosilicic acid (also called hydrofluorosilicic acid), a byproduct recovered from phosphate fertilizer manufacturing, rather than pharmaceutical-grade sodium fluoride. It can carry trace arsenic and lead. These additives must meet the NSF/ANSI Standard 60, which limits how much of each contaminant they may contribute to finished water, and measured arsenic contributions are typically a small fraction of a part per billion, well under the EPA’s 10 ppb limit (Live Science). The contamination is real but small. The stronger point is one of principle: an industrial byproduct is being added to everyone’s water to deliver a substance whose benefit is now modest.

Consent. Fluoride in water is added to treat people, not to make water safe to drink. Chlorine kills pathogens in the water; fluoride is intended to act on the body of the person drinking it. That makes fluoridation different in kind from other water treatment. A family that does not want it must buy a filter capable of removing it or buy other water, which places the burden on the person who declines rather than the one who prescribes. Utah’s governor made this point when he signed the first statewide ban in 2025, comparing fluoridation to being medicated by the government.

Claims to handle with care. Two talking points need precision. Fluoride toothpaste carries a warning to seek help if a child swallows more than the amount used for brushing; that is an acute-ingestion warning, not evidence about drinking water. And while it is true that the FDA has never approved fluoride as a drug for ingestion (in 2025 it moved against unapproved ingestible fluoride supplements for young children), fluoride added to public water is regulated by the EPA, not the FDA, so the absence of FDA approval does not by itself settle the water question.

Where the evidence is uncertain

No study has yet shown, with high confidence, that 0.7 mg/L in drinking water lowers IQ. Most studies of fluoride and IQ were done in China, India, Iran, and other places with natural fluoride well above U.S. levels, and many carry a high risk of bias. The North American pregnancy cohorts are better designed, but they are few, and their findings have been disputed on statistical grounds.

On the other side, no study has shown that 0.7 mg/L is safe for the developing brain either. For eighty years, fluoridation was justified on the assumption of a wide safety margin, and that assumption was never tested in pregnant women or infants. The burden of proof belongs to whoever proposes to add a biologically active substance to everyone’s water, not to the families who would rather not drink it.

The balance as of this writing is this. The dental benefit, measured in modern conditions, is small and uncertain. The neurological risk, at current levels, is plausible and unresolved. When a small benefit is weighed against an unresolved risk to children’s brains, and the benefit is available by another route (toothpaste) that does not require swallowing, the case for continuing to fluoridate public water is weak. Reasonable people can still disagree on how weak, and the EPA’s dose-response assessment, when it comes, should sharpen the picture.

Practical guidance for families

1. Find out what is in your water. Your utility’s annual Consumer Confidence Report lists its fluoride level and whether fluoride is added. If you are on a private well, have it tested; natural fluoride in some regions exceeds the EPA’s 2 mg/L secondary standard.
2. Protect pregnancy and infancy first. These are the windows of greatest concern. If your water is fluoridated, pregnant women may reasonably choose low-fluoride water, and infant formula should be mixed with low-fluoride water, as even the American Dental Association allows.
3. Use a filter that actually removes fluoride. Reverse osmosis, distillation, and activated alumina remove most fluoride. Standard carbon pitcher and refrigerator filters remove little or none.
4. Keep fluoride where it works: on the teeth. For children under three, use a smear of toothpaste the size of a grain of rice; for ages three to six, a pea-sized amount. Supervise brushing and teach children to spit rather than swallow. Fluoride varnish applied by a dentist is a topical option for children at high risk of decay.
5. Address the real cause of decay. Cavities are driven chiefly by frequent sugar and refined carbohydrate, especially sipped or snacked throughout the day. Limiting sugary drinks, avoiding bedtime bottles of milk or juice, and regular brushing and dental visits do more for a child’s teeth than the modest benefit fluoridated water now provides.
6. Take part in the local decision. In most states, fluoridation is decided by city councils and water boards. Utah and Florida banned it statewide in 2025, and other states are considering similar measures. Citizens can ask their local officials to review the current evidence, and they will be more persuasive if they present it accurately.

References

• National Toxicology Program. NTP Monograph on the State of the Science Concerning Fluoride Exposure and Neurodevelopment and Cognition. August 2024. ntp.niehs.nih.gov/go/fluoride
• Taylor KW, et al. Fluoride exposure and children’s IQ scores: a systematic review and meta-analysis. JAMA Pediatrics, January 2025. Summary
• Green R, et al. Association between maternal fluoride exposure during pregnancy and IQ scores in offspring in Canada. JAMA Pediatrics, 2019.
• Bashash M, et al. Prenatal fluoride exposure and cognitive outcomes in children at 4 and 6–12 years of age in Mexico. Environmental Health Perspectives, 2017.
• Iheozor-Ejiofor Z, et al. Water fluoridation for the prevention of dental caries. Cochrane Database of Systematic Reviews, October 2024. Cochrane Oral Health
• University of Manchester. Water fluoridation less effective now than in past (LOTUS study and Cochrane review). manchester.ac.uk
• U.S. EPA. Protocol for Developing the Fluoride Human Health Toxicity Assessment. EPA-822-D-26-001, August 2026. epa.gov
• Food & Water Watch v. EPA, No. 25-384 (9th Cir. May 21, 2026). National Law Review analysis; ELR summary
• Sweeney MR, et al. Fluoridation policy and practice: a European story separating myths from reality. Dublin City University, 2018. doras.dcu.ie
• FDA action on ingestible fluoride supplements for children, 2025. ADA News
• State legislative activity on fluoridation, 2026. CareQuest Institute

This article is for general information and is not individual medical advice. Readers with specific health concerns should consult their own physician or dentist.