Morning Overview

Forever chemicals turn up in nearly half of US tap water, a federal study found

Roughly 45 percent of tap water sampled across the United States contained at least one per- and polyfluoroalkyl substance, according to a U.S. Geological Survey study that tested 716 point-of-use samples drawn directly from kitchen sinks between 2016 and 2021. The findings exposed a sharp divide between households served by public water systems and those relying on private wells, raising questions about whether federal monitoring programs are catching contamination where it hits hardest.

Why the 45 percent detection rate changes the calculus for private wells

The USGS study split its 716 samples into two groups: 447 from public water supplies and 269 from private wells. Both source types showed PFAS detections, but the comparison carries a sting for the roughly 43 million Americans who depend on private wells. Public systems fall under the Safe Drinking Water Act and are subject to periodic testing through the EPA’s unregulated contaminant program. Private wells do not. That gap means a homeowner drawing water from a backyard well in a county where the nearest public utility reports clean UCMR results could still be drinking PFAS-laden water with no federal requirement to find out.

The mismatch is not hypothetical. UCMR monitoring, by design, samples only public water systems. Counties with large rural populations served primarily by private wells can appear clean in federal databases simply because no regulated utility operates there at scale. The USGS data, collected from actual kitchen taps rather than treatment-plant outlets, captured exposures that routine compliance monitoring misses. The result is a monitoring blind spot that concentrates risk on households least likely to have the resources or information to test independently.

For families on private wells, the 45 percent detection rate effectively shifts the burden of vigilance. Instead of assuming that the absence of a nearby utility violation equates to safety, well owners must treat PFAS as a plausible risk unless testing proves otherwise. In regions with known industrial sources, firefighting training sites, or historical use of PFAS-containing products, that risk is even harder to dismiss. Yet without mandatory testing, many households will not know they are exposed until a health provider, local news story, or neighbor’s test result prompts questions.

What USGS researchers found in 716 kitchen-sink samples

USGS scientists tested for 32 individual compounds across the 716 samples. At least one PFAS turned up in about 45 percent of the drinking-water samples, a figure the agency characterized as a national estimate rather than a precise prevalence rate for every community. Sampling took place directly at kitchen sinks rather than at distribution-system entry points, giving researchers a closer look at what people actually swallow after water has moved through premise plumbing and household fixtures.

Every sample in which PFOS or PFOA was detected exceeded the EPA’s 2022 interim health advisories for those two compounds, according to the USGS news release accompanying the study. That detail underscores the difference between detection and risk: even relatively low concentrations of certain PFAS can raise health concerns because advisory levels are set so close to zero. While the study was not designed to quantify health outcomes, it highlighted how widespread low-level contamination can translate into exceedances of very stringent benchmarks.

A companion USGS data release extended the work with 409 residential and commercial tap-water samples collected between May 2021 and May 2022, testing for 34 PFAS. That dataset, which covered multiple U.S. geographies including territories, used homeowner sampling kits shipped to and from households. The approach widened geographic reach but also introduced logistical limits: participation was voluntary, return rates varied, and county-level coverage has not been published in aggregated form, leaving gaps in how precisely researchers can map contamination clusters.

On the regulatory side, EPA’s current UCMR 5 cycle requires public water systems to monitor for 29 PFAS and lithium, with sample collection running from January 2023 through December 2025. Each analytical result in the UCMR 5 data system corresponds to a specific sampling location and event from a regulated public system, creating an expanding national map of PFAS occurrence in utilities large and small. The final federal PFAS drinking-water rule, issued after the USGS sampling window, established compound-specific maximum contaminant levels along with a mixture-based Hazard Index, setting enforceable limits for the first time. But those limits apply only to public systems, not to private wells.

Gaps the federal data still cannot close

Several questions remain open. The peer-reviewed analysis provides the strongest national snapshot of tap-water PFAS to date, yet 716 samples spread across a country of more than 100 million households is a thin net. Full quality-assurance records and granular site-level concentrations from the 2021–2022 data release have not been published in a form that allows independent county-by-county analysis. Without that detail, the hypothesis that private-well detections cluster in counties where public-supply monitoring shows low or no detections cannot be confirmed or rejected with the available record.

The geographic distribution of samples also matters. Urban and suburban areas with dense populations and large utilities are more likely to be represented in both USGS and UCMR datasets, while remote rural regions may have only a handful of data points, if any. That imbalance can skew risk perceptions: maps built from sparse sampling may show wide swaths of apparent “non-detects” that simply reflect a lack of testing rather than true absence of PFAS.

EPA has not issued detailed public statements explaining how UCMR 5 results, once the collection window closes at the end of 2025, will trigger enforcement actions or community notifications. The agency’s final PFAS rule sets maximum contaminant levels, but the timeline for compliance and the practical steps utilities must take to meet those limits remain subjects of ongoing implementation and, in some cases, legal challenge. Communities may see UCMR data indicating PFAS presence in their systems before formal enforcement deadlines arrive, creating pressure on local officials to act faster than regulations strictly require.

For households on private wells, the practical first step has not changed: test independently. USGS accepts questions through its public inquiry portal, and state health departments in many states offer or subsidize well-water testing kits that screen for PFAS. The cost typically runs between $200 and $400 per sample, depending on the number of compounds included and whether expedited turnaround is requested. That price is out of reach for some families, especially when multiple follow-up tests are needed to confirm results or track changes over time.

When tests do reveal PFAS, treatment options for individual well owners usually center on point-of-use filters certified to reduce PFOA, PFOS, and related compounds, or on installing whole-house systems that use activated carbon or reverse osmosis. Those systems can significantly reduce exposure but require upfront investment and ongoing maintenance. In the absence of federal mandates for private wells, state and local programs that provide financial assistance or technical guidance become the de facto safety net.

The USGS findings, combined with UCMR 5’s expanding dataset, point toward a future in which PFAS in drinking water is treated less as an isolated industrial problem and more as a diffuse national exposure issue. Yet until monitoring gaps are narrowed-especially for private wells and underserved communities-the full scope of that exposure will remain uncertain, and many households will continue to rely on incomplete information when deciding whether their tap water is safe to drink.

More from Morning Overview

*This article was researched with the help of AI, with human editors creating the final content.