On April 10 2024 the US EPA signed the first federally-enforceable Maximum Contaminant Levels for PFAS in public drinking water, 4.0 ng/L for PFOA and PFOS, 10 ng/L each for PFHxS, PFNA, and HFPO-DA (GenX), plus a unitless Hazard Index of 1 for any mixture of PFHxS, PFNA, HFPO-DA, and PFBS. The rule was published in the Federal Register on April 26 2024; community water systems must complete initial monitoring by 2027 and meet the MCLs by 2029.
The April 2024 National Primary Drinking Water Regulation (NPDWR) for per- and polyfluoroalkyl substances is the first time the US Environmental Protection Agency has set legally-binding limits on any PFAS in tap water. Before this rule, the federal guidance was non-enforceable, the 2022 interim Health Advisories for PFOA (0.004 ng/L) and PFOS (0.02 ng/L) were targets, not law. Now they are codified in 40 CFR §141.61(c) and a community water system that breaches them is in violation of the Safe Drinking Water Act.
There are six regulated compounds. Five have individual MCLs expressed in nanograms per litre, the same unit as parts per trillion. The sixth "compound" is actually a mathematical mixture rule: a Hazard Index calculated across four of the substances. The numbers below are the operative thresholds for compliance monitoring under the rule.
| Compound | Common name / abbreviation | MCL | MCLG |
|---|---|---|---|
| Perfluorooctanoic acid | PFOA | 4.0 ng/L | 0 ng/L |
| Perfluorooctane sulfonic acid | PFOS | 4.0 ng/L | 0 ng/L |
| Perfluorohexane sulfonic acid | PFHxS | 10 ng/L | 10 ng/L |
| Perfluorononanoic acid | PFNA | 10 ng/L | 10 ng/L |
| Hexafluoropropylene oxide dimer acid | HFPO-DA (GenX) | 10 ng/L | 10 ng/L |
| Mixture of PFHxS + PFNA + HFPO-DA + PFBS | Hazard Index | HI = 1 (unitless) | HI = 1 |
The unit "ng/L", nanograms per litre, is equivalent to parts per trillion (ppt). One nanogram in a litre of water is roughly one drop of ink in twenty Olympic swimming pools. The 4.0 ng/L figure for PFOA and PFOS is not approximate; the rule explicitly uses one significant figure after the decimal and that is the practical quantitation limit (PQL) the EPA determined achievable by certified laboratories using the required analytical method.
An MCL, Maximum Contaminant Level, is the legally enforceable ceiling that a public water system is allowed to deliver to its customers. It is set as close as feasible to the MCLG, taking into account analytical detection limits, treatment technology that exists in 2024, and cost. An MCLG, Maximum Contaminant Level Goal, is the non-enforceable health-based target where no known or anticipated adverse effect on human health would occur, with an adequate margin of safety. The MCLG can be below the limit of detection, even at zero, and frequently is for compounds the EPA classifies as likely human carcinogens.
For PFOA and PFOS the MCLG is zero. This reflects EPA's determination, based on the 2022 health advisory work and the underlying toxicology, that there is no safe level of either compound for lifetime exposure through drinking water. The MCL of 4.0 ng/L is therefore not a "safe" number. It is the lowest concentration that current treatment + current analytical methods can reliably enforce. For PFHxS, PFNA, and HFPO-DA the MCLG equals the MCL because the EPA's hazard assessment for those compounds did not identify them as likely carcinogens, so a non-zero threshold could be derived from reference doses without an adequate-margin-of-safety argument forcing zero.
For a deeper read on the toxicology underpinning the zero MCLG for PFOA and PFOS, see our guide on PFOA and PFOS toxicology, what the C8 cohort actually showed.
The NPDWR phases in over five years. Community water systems and non-transient non-community water systems (the same universe regulated by the Safe Drinking Water Act for chemical contaminants) must follow two clocks:
The analytical method required by the rule is EPA Method 537.1 (Rev 2.0, March 2020), or the more recent Method 533 for the shorter-chain compounds. Both use solid-phase extraction followed by liquid chromatography / tandem mass spectrometry. A laboratory must be certified by the state primacy agency to run compliance samples, and the reporting limits set by the method are what drive the 4.0 ng/L PQL for PFOA / PFOS, laboratories that cannot reliably quantify below that floor cannot be used for compliance.
For the broader monitoring background, including the EPA's Fifth Unregulated Contaminant Monitoring Rule (UCMR 5) data that fed the rulemaking, see UCMR 5 explained, what every utility had to report.
PFAS is a class of thousands of synthetic chemicals, the OECD's 2018 inventory listed over 4 700 substances meeting the structural definition, and the EPA's CompTox dashboard tracks more than 14 000. The EPA could not regulate them all at once, so it picked six on the basis of three convergent lines of evidence: detection frequency in UCMR 3 (2013–15) and the UCMR 5 monitoring beginning in 2023; the availability of a validated analytical method capable of reaching low ng/L quantitation; and a sufficient toxicological database to derive a reference dose or cancer slope factor.
The Agency for Toxic Substances and Disease Registry's 2021 Toxicological Profile for Perfluoroalkyls is the single most-cited source in the rule preamble. ATSDR's profile lays out the evidence for liver effects, immune suppression (notably reduced antibody response to childhood vaccines), thyroid disruption, developmental effects including reduced birth weight, and increased serum cholesterol. For PFOA specifically the C8 Health Project, the population study run after the 2005 DuPont settlement in the Mid-Ohio Valley, established probable links to kidney cancer, testicular cancer, ulcerative colitis, thyroid disease, pregnancy-induced hypertension, and high cholesterol. That study is summarised separately in the C8 Science Panel, what 70 000 exposed residents revealed.
PFOS sits on a comparable evidence base, dominated by the 3M Decatur cohort and post-2000 phase-out monitoring. PFHxS, PFNA, and HFPO-DA each had narrower databases but enough peer-reviewed animal and epidemiological work to support reference dose derivation. Substances with thinner data, including the short-chain PFBA and the ultra-short PFPrA, were left outside the rule for now, though the EPA flagged them for ongoing review.
The Hazard Index is the part of the rule that catches utilities off guard. Even if no single compound exceeds its individual MCL, the mixture rule can still trigger a violation. The HI applies to four compounds: PFHxS, PFNA, HFPO-DA, and PFBS (perfluorobutane sulfonic acid). PFBS is the one substance in the HI that has no individual MCL, it is regulated only through the mixture.
To compute the HI, the measured concentration of each compound is divided by its Health-Based Water Concentration (HBWC), and the four ratios are summed. The HBWCs published in the rule are: PFHxS 10 ng/L, PFNA 10 ng/L, HFPO-DA 10 ng/L, PFBS 2 000 ng/L. The formula:
HI = ([PFHxS] / 10) + ([PFNA] / 10) + ([HFPO-DA] / 10) + ([PFBS] / 2000)
If the result is greater than 1, the system is in violation regardless of individual compound levels. Worked example: a sample contains PFHxS at 6 ng/L, PFNA at 3 ng/L, HFPO-DA at 4 ng/L, and PFBS at 400 ng/L. HI = 0.6 + 0.3 + 0.4 + 0.2 = 1.5. None of the four exceeded its individual HBWC, but the mixture pushed the system over. This is the rule's recognition that PFAS rarely arrive in isolation, a contaminated source typically carries a cocktail, and joint toxicity from compounds with similar mechanisms of action is additive.
The April 2024 rule covers six compounds out of a class of thousands. Substances explicitly left out include:
Several state regulations are stricter than the federal floor, most notably the Massachusetts MassDEP rule covering a sum of six PFAS at 20 ng/L since 2020, and the New Jersey MCLs for PFNA (13 ng/L, 2018), PFOA (14 ng/L, 2020), and PFOS (13 ng/L, 2020). The April 2024 NPDWR does not preempt stricter state rules. See state PFAS regulations stricter than the EPA floor for the running map.
The EPA's regulatory impact analysis assumed three treatment technologies as compliance baselines. In practice these are also the three that water-treatment engineers have spent the last decade piloting at PFAS-affected systems:
What this means at the household level: if your utility installs GAC or IX at the entry point and operates it correctly, your tap should be below the MCLs by 2029. If your utility is far below 4.0 ng/L for PFOA / PFOS today, no action is required. If you want a faster timeline, point-of-use treatment certified to NSF/ANSI 53 (with the P473 PFOA / PFOS claim) or NSF/ANSI 58 (reverse osmosis) is the consumer-side path. The differences are covered in NSF 53 with P473 vs NSF 58, which certification actually removes PFAS and the practical comparison in reverse osmosis vs a pitcher filter for PFAS.
If you want to know where your tap sits today rather than wait for the 2027 monitoring window, the options are reviewing your utility's UCMR 5 results (free, public, but limited to the systems that sampled) or sending a sample to an accredited lab, see home PFAS testing labs that use EPA Method 537.1.
One ZIP, one PDF brief. UCMR 5 results for your state, hotspot match if applicable, and the filter that actually removes PFAS from your tap. See the hotspot map or jump to the FAQ first.
Pull my report, $15