The EPA Now Says There Aren’t 9 Million Lead Pipes—There Are 4 Million. Be Skeptical.

Just last year, the U.S. Environmental Protection Agency (EPA) published an estimate that there are nine million lead service lines (LSLs) in the nation, based on a survey of several thousand water utilities across the country. Then it reversed itself on November 25, 2025, issuing a press release discussing its allocation of funding for lead service line replacement and briefly mentioning that it has slashed its estimate of the number of lead service lines in the United States by more than half from last year’s figure to a new estimate of four million. 

Below, we analyze available information on how the EPA reached this radically different conclusion, and we raise questions about whether the new estimate is reliable. It is worth noting that previous published estimates of the number of lead pipes have ranged from 6.1 million (made by a water utility trade association in 2016) to 10 million (made by the EPA in 1991) to 9 to 12 million (originally published by NRDC in 2021 and updated last year based on our 50-state survey and industry estimates).

How did the EPA’s estimate drop from nine million to four million lead pipes?

The agency didn’t explain why the number dropped so much. But the EPA did say that it based its new extraordinarily low estimate upon new data collected from water utilities’ initial lead service line inventories, which were required to be submitted by October 2024. That sounds promising, but there are several twists. Buried deep in the EPA’s explanations are a few nuggets that could help to explain how it came up with this much lower number of LSLs:

• The EPA says, in discussing its new estimate (page 3), that “if a state did not report any data for a water system’s inventory, EPA assumed that all its service connections were non-lead.” Well, that certainly could reduce the presumed number of lead pipes. 

• The EPA admits that more than 24 million service lines are of “unknown material” in its latest estimate. Water systems reporting their inventories often just said they didn’t know the material of which some or all their pipes were made. About 6,000 community water systems said they weren’t aware of having any lead pipes. The agency guessed that less than 1 in 12 of those 24-plus million pipes of unknown material (or just 2 million) contained lead. This also probably led to a significant decrease in estimated LSLs. 

  In the past, the EPA estimated that far more of these “unknowns” were lead (see the EPA’s previous estimates of nine million lead pipes in 2023 and in 2024, based in large part on its approximations of how many unknowns were presumed to be lead, and also its 2023 Report to Congress, Appendix A). It is not clear whether the new estimate, based on utilities’ preliminary inventories, is any more accurate than last year’s, based on a survey of thousands of utilities, since both are relying on utilities’ self-reported data and extrapolating from available information to make assumptions about pipes of unknown material. But the drastic reduction raises our eyebrows.

• The EPA’s guidance (see page 2-2) to water utilities and states was to say that lead connectors to service lines (such as lead pigtails and goosenecks, which are bent lead pipes that connect the water main to the long straight pipe running up to the home) were not to be counted as LSLs. The agency said, “If the only lead pipe serving the building is a lead gooseneck, pigtail, or connector, the service line is not considered an LSL under the initial inventory requirements of the LCRR [emphasis added].” It appears from the EPA’s guidance and the templates provided to water systems that were used in developing its November 2025 estimate that the EPA did not include service lines that only had lead connectors and no other lead pipes as being LSLs. This is important because these lead connectors can still contribute to lead contamination in tap water.
• The EPA’s previous estimate and report to Congress said (page 42) that a service line that had no lead pipes but did have lead connectors such as lead pigtails or goosenecks would be considered an LSL that would qualify for replacement (LSLR) funding under the Bipartisan Infrastructure Law (BIL) State Revolving Fund (SRF) program. The agency said, “If a system reported any LSLs or connectors, EPA included those service lines in the lead content category [page 17, emphasis added].” And the agency said that “for purposes of the allotment of the BIL LSLR SRF fund, EPA considers service lines to be eligible if they contain any lead pipe or connectors or are galvanized pipe previously downstream of any lead sources [page 42, emphasis added].” From this document (see also page 20, Exhibit 2.3, Appendix A, and associated text), it appears that past EPA estimates of the number of LSLs (nine million) included service lines that only had lead connectors and no other lead pipe.

Conclusions

If the EPA had: (1) included an estimate of the number of lead pipes in the thousands of systems for which states did not report any data; (2) assumed that more than 1 in 12 service lines that the utilities said were of “unknown” material were lead, as it had previously; and (3) included service lines with lead connectors as it had earlier, the estimated number of lead service lines would be far higher than the agency’s new number of four million lead pipes. 

This is important because funding for the replacement of lead service lines under the BIL and likely under future allocations of SRF funding, depend upon the number of lead service lines in the state. The Senate Committee on Appropriations recently passed a bill (see section 453(a)) that would claw back $250 million of already appropriated FY26 BIL funding for lead service line replacement. The EPA’s reduced number of presumed lead service lines may also be a precursor to future efforts to justify cuts in funding for replacement of these lead pipes. States that have an underestimation of the number of lead pipes will receive less funding, and water utilities in those states will have a harder time getting federal funding to replace their lead pipes in the future. This is penny-wise and pound-foolish, since the health and economic benefits of removing these lead pipes are more than 14 times the costs. And it does not bode well for the tens of millions of Americans who continue to drink lead-contaminated water from these lead pipes, which have been linked to harm to children’s brains and adults’ cardiovascular health, among other serious health harms.