National analysis
PFAS in US tap water: what EPA's own testing found, state by state
EPA finished the largest PFAS testing programme in American history. We processed every result and found detections in a third of systems tested, serving roughly 152 million people.
The short answer: EPA's fifth Unregulated Contaminant Monitoring Rule tested 10,297 community water systems for 29 PFAS compounds. We found detections in 3,539 of them, or 34 percent. Those systems serve about 152 million people, roughly half the population covered by the programme. Detection rates vary enormously by state, from 68 percent of systems in New Jersey down to 4 percent in Mississippi.
Where the numbers come from
UCMR 5 is the closest thing America has to a national PFAS census. Between 2023 and 2025, EPA required every water system serving more than 3,300 people, plus a sample of smaller ones, to test for 29 individual PFAS compounds at reporting limits in the low parts per trillion. The results are public, machine readable, and largely unread.
We downloaded the full occurrence dataset and processed it ourselves. That processing matters: the standard pipeline for reading these files contains a units-handling bug that silently discards detections. EPA reports results in micrograms per litre using a micro sign, and when that character is mangled by an encoding mismatch, naive converters drop the value and mark it uncertain. Run that way, the national dataset appears to show zero PFAS detections in the United States. We caught it because zero is obviously wrong, fixed the conversion, and recovered 3,539 systems with genuine detections.
Which states have the most PFAS in their water?
Among states with at least 40 systems tested, detection rates run like this:
| State | Systems with a detection | Systems tested | Rate |
|---|---|---|---|
| New Jersey | 164 | 241 | 68.0% |
| Connecticut | 42 | 63 | 66.7% |
| South Carolina | 91 | 164 | 55.5% |
| Massachusetts | 143 | 262 | 54.6% |
| Texas | 601 | 1,121 | 53.6% |
| North Carolina | 139 | 282 | 49.3% |
| Florida | 185 | 388 | 47.7% |
| Pennsylvania | 159 | 360 | 44.2% |
| Montana | 5 | 40 | 12.5% |
| South Dakota | 6 | 55 | 10.9% |
| Arkansas | 10 | 171 | 5.8% |
| Hawaii | 2 | 44 | 4.5% |
| Mississippi | 9 | 224 | 4.0% |
Read that spread carefully, because it is partly a story about industry and partly a story about geology. Dense industrial history, firefighting foam use at airports and military bases, and heavy reliance on surface water all push rates up. States drawing from deep, protected groundwater tend to sit lower.
Which large water systems had the highest results?
Among systems serving 100,000 people or more, these recorded the highest single PFAS result in UCMR 5:
| System | State | Population served | Highest single result |
|---|---|---|---|
| Emerald Coast Utilities Authority | FL | 242,172 | 220 ppt |
| Miami-Dade WASD main system | FL | 2,377,460 | 190 ppt |
| GWA Northern Water System | GU | 139,207 | 148 ppt |
| Artesian Water Company | DE | 231,114 | 106 ppt |
| City of League City | TX | 149,557 | 102 ppt |
| City of Hialeah | FL | 238,000 | 91.1 ppt |
| City of Grand Prairie | TX | 201,843 | 82 ppt |
For scale: EPA's enforceable limits for PFOA and PFOS are 4 parts per trillion each, phased in through 2029. Several compounds appearing in these results, including 6:2 FTS and PFPeA, have no federal limit at all yet. The highest single result anywhere in the dataset was 1,700 ppt at Saint Paul Park, Minnesota, a small system near a long-running industrial site.
Against those limits, 1,504 systems serving 69.8 million people already measure above them, ahead of the 2029 compliance deadline.
Which PFAS compounds show up most often?
The compounds people have heard of are not the ones most frequently detected. Across all systems with a detection, the leaders were PFPeA (1,824 systems), PFBA (1,676), PFHxA (1,424), PFBS (1,180), PFOS (1,167) and PFOA (1,041). PFOA and PFOS get the headlines and the regulation because their health effects are best characterised, but the shorter-chain compounds are more widespread, and most of them are unregulated.
Does a detection mean the water is unsafe?
No, and this is where most coverage goes wrong. A detection means the lab found a compound above its reporting limit, which for these methods is often below 2 ppt. Many detections sit well under any health-based guidance. What a detection does tell you is that PFAS reached your source water, that the level is worth knowing, and that a non-detect elsewhere is not luck but a genuinely different situation.
The reverse error is more dangerous. A system with no UCMR 5 record has not been shown to be clean; it has not been tested. We never label an untested system as PFAS-free, and you should be suspicious of any site that does.
What can you actually do?
Look up your own system first, because state averages tell you very little about your tap. If your system had a detection and you want to act, the filter standard that matters is NSF/ANSI 53 certified specifically for PFOA and PFOS reduction, or NSF/ANSI 58 reverse osmosis, which handles a broader range. An ordinary carbon pitcher certified only for taste and chlorine does essentially nothing for PFAS, and the packaging rarely makes that obvious.
Common questions
How many US water systems have PFAS in them?
Which state has the worst PFAS contamination in drinking water?
Is 190 ppt of PFAS dangerous?
Does boiling water remove PFAS?
See what your own utility reported
Fifty of the largest US systems, scored against health guidance, every figure linked to its source.