Pregnancy & fertility

Fluoride during pregnancy: what the research actually says

The studies that drove the headlines used fluoride levels 2–14 times higher than US tap water. Here is what the evidence does — and doesn't — say about 0.7 mg/L.

WH

WaterHealthCheck Editorial

Reviewed for accuracy against peer-reviewed literature and EPA guidelines · 9 min read · Updated May 2026

The short answer

At the level in US municipal water — 0.7 mg/L — fluoride is not a proven harm to pregnancy. The research that raised concern comes from areas where naturally occurring fluoride reaches 1.5–4+ mg/L, not the 0.7 mg/L in fluoridated US tap water. There is genuine scientific uncertainty at levels closer to the US municipal threshold, introduced by a 2022 NTP systematic review — but it is not settled science, and official bodies including the CDC, AAP, and ADA have not changed their guidance. If you want to reduce fluoride exposure as a precaution, reverse osmosis removes it effectively. This is a reasonable personal decision, not an emergency.

What is fluoride and why is it in tap water?

Fluoride is a naturally occurring mineral found in soil, rock, and water. In the US, community water systems add fluoride to a target concentration of 0.7 mg/L — a level set by the US Public Health Service in 2015 specifically to preserve the dental health benefits of fluoridation while reducing the risk of dental fluorosis (cosmetic mottling of tooth enamel from excess fluoride).

Not all US water is fluoridated: some utilities serve naturally fluoridated groundwater, others serve wells with no fluoride at all. Parts of the Southwest — particularly sections of Texas, New Mexico, and Arizona — have naturally occurring fluoride levels that can exceed 1 mg/L in groundwater, occasionally significantly higher in some private well areas.

Fluoride levels at a glance

0.7 mg/L
US municipal fluoridation targetThe standard level in fluoridated community water systems
1.0 mg/L
EWG health guidelineClose to — but above — the US fluoridation level
2.0 mg/L
EPA secondary MCLNon-enforceable aesthetic guideline (taste / appearance)
4.0 mg/L
EPA primary MCL (enforceable)Legal maximum — exceeding this requires public notification
1.5 – 4+ mg/L
Levels in concerning researchRange used in the Canadian MIREC study and Chinese studies cited in media coverage

What does the research actually say?

The fluoride-and-pregnancy literature is genuinely mixed, and understanding what the studies measured matters more than reading the headlines about them.

The studies that raised concern

Green et al. (2020) — the Canadian MIREC study

Widely cited

This prospective cohort study of 512 Canadian mother-child pairs found an association between higher maternal urinary fluoride during pregnancy and lower IQ scores in boys at ages 3–4 — specifically a 4.49-point reduction in boys' IQ per 1 mg/L increase in maternal urinary fluoride. This study received significant media attention and is frequently cited as evidence that fluoride is dangerous in pregnancy.

The important caveats

  • The Canadian study population lived in areas with fluoride ranging from 0.04 to 1.1 mg/L — not dramatically higher than US levels, but the association was dose-dependent across that range.
  • The effect was only statistically significant in boys, not girls — a sex-specific finding that has not been consistently replicated and raises questions about confounding.
  • The study measured urinary fluoride (total body exposure), not drinking water fluoride. Fluoride intake from toothpaste, food, and beverages other than water is not captured in utility testing data.
  • It is an association study — it cannot establish causation. Confounding by socioeconomic factors, co-exposures, and dietary differences is a recognised limitation.
  • Multiple research groups have attempted to replicate the association in US populations and have not found it consistently.

NTP systematic review (2022 / published 2024)

Policy-relevant

The National Toxicology Program conducted a systematic review of fluoride's neurodevelopmental effects across 74 human studies and classified fluoride as a "presumed" cognitive neurodevelopmental hazard based on a consistent pattern of lower IQ in children exposed to higher fluoride levels. This was a significant finding from a rigorous federal review body and it did receive serious scientific attention.

What the NTP review covered — and what it didn't conclude

  • The majority of studies reviewed were from China, India, Iran, and other countries where naturally occurring fluoride ranges from 1.5 to 10+ mg/L — far above the US 0.7 mg/L fluoridation level.
  • The NTP review specifically noted that most of the studies it evaluated used fluoride concentrations higher than those in US-fluoridated water systems.
  • The review did not conclude that 0.7 mg/L is harmful. It classified fluoride as a presumed hazard at levels associated with concern — a broader statement covering a range of exposures.
  • The NTP review did not lead to a change in public health policy by the CDC, AAP, or ADA.
  • CDC, AAP, and ADA continued to support community water fluoridation at 0.7 mg/L as safe and beneficial after the review was published.

Studies finding no concern at US levels

Several studies conducted in the US specifically — where fluoridation is at 0.7 mg/L — have not found associations between community fluoridation and developmental or IQ outcomes in children. A large 2021 study (Broadbent et al.) found no association between fluoride exposure and IQ in a New Zealand cohort followed through adulthood. Multiple US-based studies using UCMR and water system data have similarly failed to replicate the association found in the Canadian study.

The honest picture is this: there is a consistent signal of neurodevelopmental harm at fluoride levels of 1.5 mg/L and above. That signal becomes less consistent and less clear at levels closer to 0.7 mg/L. Whether the effect extends to 0.7 mg/L is genuinely uncertain — not "proven safe" but also not "proven harmful." This is the state of the science as of 2026.

What the internet gets wrong about fluoride and pregnancy

The majority of alarming claims about fluoride in pregnancy fall into one of two categories:

Citing Chinese studies with 2–10 mg/L fluoride

A large body of research from China and India describes neurodevelopmental harm in areas with severe endemic fluorosis — regions where naturally occurring groundwater fluoride reaches 2 to 10+ mg/L. These are areas where skeletal fluorosis is visible in the population. Applying these findings to US municipal water at 0.7 mg/L is a category error. The exposures are categorically different.

Treating the NTP review as a ban recommendation

The NTP systematic review is the most credible piece of recent evidence for concern about fluoride at population levels — but it classified fluoride as a "presumed" hazard across a range of exposures, not as harmful specifically at 0.7 mg/L. Federal agencies reviewed it and did not change public health guidance. Describing it as proof that US tap water is unsafe in pregnancy is not an accurate reading of what the review concluded.

What is accurate

The EWG health guideline of 1.0 mg/L sits just above the 0.7 mg/L US fluoridation level — a genuinely narrow margin. The scientific uncertainty introduced by the NTP review is real, even if its implications for 0.7 mg/L specifically are not settled. Choosing to reduce fluoride exposure as a precaution during pregnancy is a defensible personal decision. It just is not required by the weight of current evidence.

Should you filter fluoride during pregnancy?

The answer depends on your utility's actual fluoride level — and your personal threshold for precaution given genuine scientific uncertainty.

Your utility is at the standard 0.7 mg/L

The evidence does not require filtration. This is a personal decision. If the uncertainty from the NTP review concerns you, filtering is a reasonable precaution with no downside.

Your utility reports fluoride above 1.0 mg/L

Precaution warranted

You are above the EWG health guideline. Reducing exposure during pregnancy is a reasonable precaution at this level, regardless of the unresolved debate about 0.7 mg/L.

You are in the Southwest, Texas, or New Mexico on well water

Check your level

Naturally occurring fluoride in some areas of these states reaches 1.5–4+ mg/L. Check your actual water test result — this is the scenario where the concerning research is most applicable.

Your utility is fluoride-free or below 0.5 mg/L

Fluoride is not a meaningful concern for your specific water. Focus on other contaminants — lead, TTHMs, nitrate, and PFAS are the ones with documented pregnancy risk.

Check your utility's fluoride level

The national debate about fluoride is less useful than knowing your specific utility's reported level. Enter your ZIP code to see your utility's data alongside other contaminants with documented pregnancy risk.

Free · No email required · Powered by EPA SDWIS + UCMR5 data

How to remove fluoride from drinking water

If you decide to reduce fluoride exposure during pregnancy, these are the methods that actually work — and the ones that don't.

Reverse osmosis (NSF 58 certified)

~90–95% fluoride removal

The most comprehensive option. Removes fluoride, nitrate, PFAS, and most other dissolved contaminants. Under-sink systems cost $200–500; countertop options are available. Uses a semi-permeable membrane to filter dissolved minerals.

Distillation

Effectively complete fluoride removal

Boiling water and collecting steam removes fluoride and other dissolved minerals. Countertop distillers are available at $150–300. Slower than RO and produces smaller volumes, but effective.

Activated alumina filter (NSF 53 certified for fluoride)

~90–95% fluoride removal

A specific filter media certified for fluoride reduction. Less common than RO, but effective. Verify the specific NSF 53 certification lists fluoride — not all NSF 53 filters include it.

Standard carbon filters (Brita, PUR, most pitcher filters)

Does not remove fluoride

Activated carbon is excellent at reducing chlorine, TTHMs, taste, and some heavy metals — but it does not remove fluoride. NSF 42 certified filters address aesthetic issues only. If your concern is fluoride specifically, a pitcher filter will not help.

Boiling water

Makes fluoride worse

Boiling removes water through evaporation, concentrating the fluoride that remains. Do not boil water to remove fluoride.

A practical note: if you are also concerned about TTHMs, nitrate, or PFAS — the contaminants with stronger and better-replicated evidence for pregnancy risk — reverse osmosis addresses all of them at once. It is the most comprehensive filter option for pregnancy, covering fluoride alongside the contaminants with more established risk profiles.

Fluoride and infant formula / breastfeeding

Two separate questions are often conflated here, and they have different answers.

Breastfeeding

Fluoride does not transfer meaningfully to breast milk regardless of how much fluoride a mother consumes. Human breast milk contains very low fluoride concentrations (typically around 0.01–0.02 mg/L) that are largely independent of maternal intake. There is a physiological barrier that limits fluoride transfer to milk. From a breastfeeding standpoint, your water's fluoride level is not a concern for the infant.

Infant formula preparation

This is where fluoride does matter — not for toxicity, but for the cosmetic risk of dental fluorosis. When formula is reconstituted with fluoridated water, infants receiving formula as their primary nutrition can ingest more fluoride than recommended, potentially increasing the risk of mild enamel fluorosis (white spots or faint streaking on permanent teeth, which erupt years later).

Dental fluorosis is a cosmetic issue, not a health risk. However, the American Dental Association recommends that parents using powdered or liquid concentrate formula occasionally substitute low-fluoride water (bottled water labeled as purified, deionised, or demineralised, or water processed through reverse osmosis) when preparing formula — particularly during the first year of life when infant formula is the primary food source.

Ready-to-feed formula (pre-mixed liquid) does not require water preparation and sidesteps this issue entirely.

How to think about this decision

The fluoride-and-pregnancy question exists on a spectrum from "no concern" to "unresolved uncertainty" — not from "proven safe" to "proven dangerous." Here is a clear framework for making the decision:

  1. 01

    Find out your utility's actual fluoride level

    Not the national average. Not what you read about. Your specific utility's data. The tool on this page pulls EPA monitoring data for your ZIP code. This is the relevant number.

  2. 02

    Compare to the reference levels

    0.7 mg/L: standard US fluoridation. 1.0 mg/L: EWG health guideline. 1.5+ mg/L: the threshold where most concerning research was conducted. If your utility is at 0.7 mg/L, the debate about higher-level studies is not directly applicable to your water.

  3. 03

    Make a proportionate decision

    At 0.7 mg/L: filtering is a reasonable precaution if the NTP review's uncertainty concerns you, but not required by evidence. Above 1.0 mg/L: precaution is more clearly warranted. Above 2.0 mg/L: you are above the EPA secondary MCL and should definitely consider filtration.

  4. 04

    Keep fluoride in proportion to other water quality concerns

    While the fluoride debate has attracted significant media attention, TTHMs, lead, nitrate, and PFAS have stronger and better-replicated evidence for pregnancy risk at levels found in US tap water. If you are going to install a filter, prioritise one that addresses those contaminants — and if it removes fluoride too (as reverse osmosis does), that is a bonus.

  5. 05

    Mention it to your provider if it concerns you

    Your OB or midwife may have perspective on local water quality and can factor your specific situation into their guidance. Your ZIP code report is a useful starting point for that conversation.

Sources and methodology

  • US Public Health Service. U.S. Public Health Service Recommendation for Fluoride Concentration in Drinking Water for the Prevention of Dental Caries (2015). Public Health Reports.
  • Green R, et al. (2020). Association between maternal fluoride exposure during pregnancy and IQ scores in offspring in Canada. JAMA Pediatrics.
  • National Toxicology Program. Systematic Review of Fluoride Exposure and Neurodevelopmental and Cognitive Health Effects (2024). NTP Monograph. ntp.niehs.nih.gov
  • Broadbent JM, et al. (2015). Community water fluoridation and intelligence: Prospective study in New Zealand. Am J Public Health.
  • EPA. National Primary Drinking Water Regulations — Fluoride. epa.gov/ground-water-and-drinking-water/national-primary-drinking-water-regulations
  • EWG Tap Water Database — Fluoride health guideline. ewg.org/tapwater
  • American Dental Association. Fluoridation FAQs. ada.org
  • CDC. Community Water Fluoridation — Frequently Asked Questions. cdc.gov/fluoridation
  • American Academy of Pediatrics. Fluoride Use in Caries Prevention in the Primary Care Setting (2014). Pediatrics.
  • Iheozor-Ejiofor Z, et al. (2015). Water fluoridation for the prevention of tooth decay. Cochrane Database of Systematic Reviews.
  • Till C, et al. (2020). Association of fluoride exposure with cardiometabolic stress markers and sleep in the Canadian MIREC cohort. Environment International.
  • Saxena S, et al. (2012). Fluoride accumulation in the placenta. Fluoride (journal of the International Society for Fluoride Research).