PFAS statistics span a broad chemical class, so water concentrations, detections, populations served, and study-level findings must be read with their definitions and measurement periods. The figures below distinguish observed monitoring from modeled estimates and preserve the limits of each dataset.
Key PFAS Statistics
The most broadly useful PFAS figures include:
- 71–95 million people in the Lower 48 United States may rely on groundwater with detectable PFAS in the USGS 2023 study release.
- More than 20% of the Lower 48 population may rely on PFAS-detectable groundwater according to the USGS 2023 modeled comparison.
- 4,000–6,500 U.S. public water systems could exceed PFOS or PFOA maximum contaminant levels in EPA’s 2024 national occurrence estimate.
- 82–103 million people could be served by U.S. systems exceeding PFOS or PFOA maximum contaminant levels in EPA’s 2024 estimate.
- 4,100–6,700 U.S. public water systems could exceed at least one applicable PFAS limit in EPA’s combined 2024 estimate.
- 82–105 million people could be served by systems exceeding at least one applicable PFAS limit in EPA’s combined 2024 estimate.
- PFOA appeared at or above 20 ng/L in 1.03% of UCMR 3 finished-water samples collected in the United States from 2013–2015.
- PFOA was detected in 117 of 4,920 monitored U.S. public water systems during UCMR 3 from 2013–2015.
- PFOA detections in UCMR 3 ranged from 20 to 349 ng/L in U.S. finished water from 2013–2015.
- PFNA appeared in 0.05% of UCMR 3 samples at or above 20 ng/L in U.S. finished water from 2013–2015.
- PFBS appeared in 0.05% of UCMR 3 samples at or above its 90 ng/L reporting level in U.S. finished water from 2013–2015.
- 254 groundwater samples were collected in 2019 across five eastern U.S. aquifer systems for a USGS study.
- A global groundwater meta-analysis used 96 reports from 1999–2021 covering 20 countries.
- Approximately 21,000 groundwater data points remained after exclusions in that global meta-analysis covering 1999–2021 literature.
- A global review compiled 16,246 groundwater data points for 24 priority PFAS from studies published since 2001.
- A global review compiled 77,541 surface-water data points for 24 priority PFAS from studies published since 2001.
- PFOA was detected in 102 of 122 global drinking-water studies in a review of literature from 2000 through February 2025.
- PFOS was detected in 100 of 122 global drinking-water studies in the same 2000–February 2025 review.
- UCMR 5 requires monitoring for 29 PFAS plus lithium during U.S. sample collection from January 2023 through December 2025.
Contents
- PFAS statistics in U.S. drinking-water systems
- PFAS detection statistics by chemical and concentration
- PFAS groundwater and source-water data
- State-level PFAS drinking-water trends
- Global PFAS occurrence statistics in water
- PFAS monitoring scale, health evidence, and regulatory estimates
PFAS Statistics in U.S. Drinking-Water Systems
EPA’s 2024 national occurrence estimates describe systems and people served, not individually measured exposure. The estimates are modeled national extrapolations, and the categories overlap where one system could exceed more than one limit.
| EPA estimate, 2024 | Lower bound | Upper bound |
|---|---|---|
| Systems exceeding PFOS or PFOA MCLs | 4,000 | 6,500 |
| People served by systems exceeding PFOS or PFOA MCLs | 82 million | 103 million |
| Systems exceeding the PFAS hazard-index MCL | 300 | 700 |
| People served by hazard-index systems | 9 million | 18 million |
| Systems exceeding at least one applicable limit | 4,100 | 6,700 |
| People served by systems exceeding at least one limit | 82 million | 105 million |
Source: EPA, Per- and Polyfluoroalkyl Substances Occurrence and Contaminant Background Support Document.
The combined rows should not be created by adding the separate rows. EPA’s combined modeled range already accounts for overlap among PFOS, PFOA, and hazard-index categories.
The USGS provides a different population perspective. Its 2023 study release estimated that 71–95 million people in the Lower 48 may rely on groundwater with detectable PFAS for drinking-water supplies, representing more than 20% of the Lower 48 population.
That USGS result is a model-based estimate of detectable concentrations and groundwater reliance, not a direct census of contaminated taps or individual exposure.
PFAS Detection Statistics by Chemical and Concentration
UCMR 3 collected 36,972 finished-water samples from 4,920 public water systems during 2013–2015. A detection means a result at or above the chemical-specific minimum reporting level (MRL), so below-MRL results should not be read as zero.
| Chemical | UCMR 3 detection measure | Detected range |
|---|---|---|
| PFOA | 1.03% of samples; 117 systems, or 2.4%; systems served 3.2% of monitored population | 20–349 ng/L |
| PFNA | 19 samples, or 0.05%; 14 systems, or 0.28%; systems served 0.22% of population | 22–55.88 ng/L |
| PFBS | 0.05% of samples; 8 systems, or 0.16%; systems served 0.15% of population | 90–370 ng/L |
The PFOA MRL was 20 ng/L. PFNA also used a 20 ng/L MRL, while PFBS used a 90 ng/L MRL; the system and population percentages refer to at least one result at or above the applicable MRL.
Source: EPA, PFAS occurrence support document.
State-reported finished-water data summarized by EPA through 2023 show why chemical and threshold definitions matter. PFOA detections ranged from 0.21 to 650 ppt, PFHxS from 0.2 to 856 ppt, PFNA reached 330 ppt at the upper end, and PFBS ranged from 0.22 to 720 ppt.
For water, 1 ppt equals 1 ng/L. These state datasets used varying reporting thresholds and do not form a probability distribution for all U.S. systems.
PFAS Groundwater and Source-Water Data
The eastern-U.S. USGS study sampled groundwater used as a drinking-water source in 2019, collecting 254 samples across five aquifer systems and analyzing 24 PFAS compounds. Its predictive analysis evaluated 57 chemical and land-use variables as possible predictors of detection, but predictor importance is model output rather than causal proof.
Source: USGS, PFAS in groundwater used as a source of drinking water in the eastern United States.
Global groundwater evidence is assembled from published studies rather than a uniform monitoring network. A meta-analysis covering 96 reports published from 1999 through 2021 across 20 countries included more than 35 PFAS compounds.
After removing time-series and duplicate samples and excluding non-detects, that meta-analysis retained approximately 21,000 groundwater data points. Because non-detects were excluded, the total is not a prevalence estimate.
Source: Global distributions, source-type dependencies, and concentration ranges of PFAS in groundwater.
Another global occurrence review compiled 16,246 groundwater data points and 77,541 surface-water data points for 24 priority PFAS from studies published since 2001. The review notes that study locations and analytical detection limits varied, so the totals describe research coverage rather than a uniform global survey.
Source: PFAS in water and wastewater: A critical review of global occurrence and distribution.
State-Level PFAS Drinking-Water Trends and Differences
EPA’s state tables combine different periods, source-water mixes, reporting thresholds, and coverage. The comparisons below are descriptive, not direct rankings.
| Geography and chemical | Population or sample result | Threshold or period |
|---|---|---|
| Arizona groundwater, PFOS | 50.0% of 24 samples detected; 25.0% exceeded 10 ppt | Arizona monitoring period summarized by EPA |
| Luke Air Force Base groundwater, PFOS | 28.8% of 264 samples detected; 10.6% exceeded 10 ppt | Site-specific Arizona data |
| Illinois groundwater, PFHxS | 14.2% of 1,823 samples detected; 0.9% exceeded 10 ppt | Illinois monitoring period |
| Illinois surface water, PFHxS | 8.9% of 302 samples detected; none exceeded 10 ppt | Illinois monitoring period |
The Arizona state reporting threshold was 1.6–2 ppt, while Illinois used 1.7–3.7 ppt for the PFHxS data. The Luke Air Force Base result is site-specific and should not be treated as representative of Arizona overall.
Additional population-served results show large descriptive differences:
- In New Hampshire, 56.3% of people served by groundwater systems were served by systems with PFOA detections, and 16.1% by systems above 10 ppt; for surface-water systems, the corresponding figures were 31.1% and 1.3%.
- In New Jersey, 47.7% of people served by groundwater systems were served by systems with PFOA detections, and 25.5% by systems above 10 ppt.
- In Virginia, 41.3% of people served by surface-water systems were served by systems with PFOA detections, and 34.9% by systems above 5 ppt; no systems exceeded 10 ppt in the cited table.
- In Wisconsin during 2022–April 2023, detections affected 3.6% of the population served by groundwater systems and 20.5% of the population served by surface-water systems.
- In Vermont during 2019–April 2023, 1.8% of the covered population was served by systems with PFOA detections and 0.3% by systems above 10 ppt; the covered population was 385,830.
Source for the state results: EPA, PFAS occurrence support document.
New Hampshire reporting thresholds varied from 2–5 ppt, New Jersey thresholds from 0.018–8.9 ppt, and Virginia’s reporting threshold was 3.5 ppt. Wisconsin sampling was ongoing and its reporting thresholds were not reported; Vermont sampling was also ongoing.
Global PFAS Occurrence Statistics in Water
A 2025 global drinking-water review included 122 studies from literature published from 2000 through February 2025. It found PFOA in 102 studies, or 64.69% of studies, and PFOS in 100 studies, or 60.72% of studies.
These are study-level detection percentages, not the percentage of all drinking-water samples, systems, or people worldwide. The underlying studies differed in geography, methods, and sampling design.
Research attention also has its own measurable distribution. A systematic review of total-PFAS analytical methods found that water was the sample type in 34% of included articles during the review’s publication period.
Source: A systematic review of methods for the analysis of total PFAS.
That 34% figure describes the share of research articles using water samples, not environmental prevalence or the share of contaminated water supplies.
PFAS Monitoring Scale, Health Evidence, and Regulatory Estimates
U.S. monitoring has expanded beyond the earlier UCMR 3 snapshot. UCMR 5 requires monitoring for 29 PFAS plus lithium during sample collection from January 2023 through December 2025.
Source: EPA, Occurrence Data from the Unregulated Contaminant Monitoring Rule.
For its 2024 national occurrence model, EPA used 65,537 samples from 28 state datasets in addition to UCMR 3 data. The inputs supplemented UCMR 3 and were not a single uniform survey.
EPA also reported that PFNA and PFBS each had 19 above-MRL results among nearly 37,000 samples per chemical in UCMR 3. Sparse above-MRL counts do not describe concentrations below the reporting level.
Source: EPA, PFAS occurrence support document.
Health-related measurements in the supplied evidence are local biomonitoring observations rather than national estimates. A peer-reviewed review reported PFOA blood concentrations of 0.4–7.3 μg/L, with a median of 1.75 μg/L, among residents near a fluorochemical plant in North Carolina.
The same review reported PFOS blood concentrations of 1.4–34.6 μg/L, with a median of 5.4 μg/L, among residents near that plant. These findings describe a local exposed community and are not representative of the general population.
Source: PFAS in water and wastewater: A critical review of global occurrence and distribution.