Members of Wisconsin’s high school class of 1957 who grew up with fluoridated drinking water scored no lower on IQ tests as teenagers than classmates who grew up without it, once family background was taken into account. Decades later, on thinking tests given at ages 53 to 80, there was still no meaningful difference. The finding comes from a study of 10,317 people published in April 2026 in the Proceedings of the National Academy of Sciences.

The question has practical weight. Utah and Florida ended community water fluoridation in 2025, and a number of U.S. cities and counties have moved the same way, often citing worries about children’s brain development. This is one study, built on one state’s records from the 1940s and 1950s, and it is not the final word. Nor is this article medical or dental advice; it is a report on what the research does and does not show.

What the Wisconsin study measured

The paper, a brief report led by sociologist John Robert Warren of the University of Minnesota with colleagues at the University of Wisconsin and the University of Michigan, draws on the Wisconsin Longitudinal Study. That project took a random one-third sample of the state’s 1957 graduating class and has followed the same people ever since.

The researchers sorted participants into four groups. About 3,600 had no fluoridated water in childhood. About 2,600 were counted as exposed from birth, because untreated wells in their county carried naturally sufficient fluoride. The remaining 4,100 or so lived in communities that began fluoridating partway through childhood, from about age 8 or from about age 14.

IQ came from the Henmon-Nelson test, which schools gave in students’ freshman and junior years. Later thinking was measured with a reasoning task from the Wechsler Adult Intelligence Scale at ages 53, 64 and 72, and with a telephone test of cognitive status at 80.

After adjusting for family income, parents’ education, father’s occupation, community size and school characteristics, none of the exposed groups scored significantly lower, or higher, than the unexposed group at any age. Of 45 key comparisons, only two reached statistical significance, roughly what chance alone would produce.

Because more than 90 percent of the records had been linked to the 1940 and 1950 censuses, the team could also repeat the analysis using only people who had lived in the same county since age 11, and then only those who had stayed put since age 1. Those are the participants whose exposure is least likely to have been misclassified. The result held.

Why the raw numbers flattered fluoride

Before adjustment, the group whose towns began fluoridating when they were about 14 had the highest average scores at every age. That group also had the highest family incomes and came from the largest communities. The never-exposed group came from the smallest communities and the lowest incomes.

Once family and school background were accounted for, the advantage disappeared.

This is the central difficulty in fluoride research, and it runs in both directions. A naive comparison of towns in 1950s Wisconsin would have made fluoride look good for the brain. Studies that find lower IQ in high-fluoride regions face the mirror-image problem, since naturally high fluoride often occurs alongside rural poverty and other exposures that also affect children.

Warren’s group reached a similar conclusion in an earlier study in Science Advances in November 2025. Using the nationally representative High School and Beyond survey of U.S. students first questioned in 1980, they found children exposed to recommended fluoride levels scored modestly better on school achievement tests, an edge that was smaller and no longer statistically significant by about age 60. That study used achievement tests rather than IQ, which is the gap the new paper set out to fill.

The evidence that pointed the other way

Much of the recent concern traces to a 2025 meta-analysis in JAMA Pediatrics led by Kyla Taylor of the U.S. National Institutes of Health. It pooled 74 studies and found higher fluoride exposure was associated with lower IQ in children. Science Blog reported on that review when it appeared.

The details matter. Of the 74 studies, 45 were from China and 12 from India, and most were judged to be at high risk of bias. The clearest associations appeared at fluoride levels above 1.5 milligrams per liter, the World Health Organization’s guideline value. For drinking water below that level, the authors described the evidence as limited and uncertain. The current U.S. recommendation for fluoridated water is 0.7 milligrams per liter.

In the new paper, the Wisconsin authors point out that none of the evidence in that review came from population-representative samples or from data collected in the United States. That is their reading. Others, including Bruce Lanphear, who wrote an editorial accompanying the meta-analysis, have argued that health agencies should still reassess fluoride’s risks, particularly for pregnant women and infants.

What the study cannot show

Christine Till, a neuropsychologist at York University in Toronto, told Science News that the results “should be interpreted cautiously.” The class of 1957 was born before fluoridation was widespread, so for most participants the study says little about exposure in the womb or in infancy, periods she describes as key for brain development. Steven Levy, a dental public health researcher at the University of Iowa who was not involved, called it “very strong data” in the same report.

Exposure was inferred from where people lived. Nobody’s fluoride intake was measured, and the study had no information on dental care or on fluoride from sources other than water. The authors list both as limitations.

The main paper does not give the fluoride concentrations in participants’ water, so its result cannot be pinned to a specific dose such as today’s U.S. recommendation.

Two further cautions apply. Both U.S. studies finding no harm come from the same research group, so they do not independently confirm each other. And Wisconsin teenagers of the 1950s, whose other fluoride sources went unrecorded, are an imperfect stand-in for children born today.

A question that reaches Australia

I’m interested in this because here in Australia, there’s fluoride in the water supply of Melbourne, Victoria, but not in Cairns. So I’m curious about it.

Melbourne has had fluoridated water since 1977, according to Victoria’s Department of Health, which sets a maximum of 1 milligram per liter. Cairns Regional Council voted to remove fluoride in 2013, after Queensland made fluoridation a decision for local councils in 2012. In December 2024 the council decided not to reintroduce it.

Australia has some direct evidence of its own. A study led by Loc Do of the University of Queensland, published in the Journal of Dental Research in December 2024, gave full Wechsler IQ tests to 357 young Australians aged 16 to 26 whose residential histories had been recorded in a national child oral health survey about a decade earlier. Those who had fluoridated water for all of their first five years scored about one point higher than those who had none, a difference too small to distinguish from zero. The authors concluded the scores were equivalent. That study covered early childhood, the period the Wisconsin cohort largely missed, though its sample was far smaller.

Anyone weighing fluoride for their own family, including toothpaste or supplements for young children, is best placed to talk it over with a dentist or doctor.

The open question is whether fluoride at today’s levels has any effect before birth or in the first years of life that these studies were too coarse to detect. Answering it would take a large study that measures each person’s actual fluoride intake, starting in pregnancy, and follows the children into adulthood.