What Best Available Data Tells Us About Lead Service Lines
By Maria Sevillano & Erica Galante-Johnson
In November 2025, EPA released the first national dataset based on utility-reported lead service line (LSL) inventories. Under the Lead and Copper Rule Revisions (LCRR) and subsequent Lead and Copper Rule Improvements (LCRI), water systems were required to identify and report service line materials by October 2025 and update that information annually.
Critically, EPA has now tied federal lead service line replacement funding to inventory data rather than relying on a survey of utilities that EPA runs (DWINSA). We have always supported this shift because it will provide a more accurate picture of replacement needs and better aligns the allocation of federal funds with actual state conditions.
This blog examines what the new inventory data tells us, how it compares to prior survey-based estimates, and what it reveals about progress and remaining gaps. The short version: we now know far more about where lead pipes remain; some states are close to eliminating them, but data quality still matters.
Surveys vs Inventories: Why the Difference Matters
The 7th DWINSA, released in April 2023, estimated LSL replacement needs using a representative sample of water systems. It was designed for high-level planning purposes, not for precise allocation of funds. By contrast, the LCRR inventory is a comprehensive, system-by-system reporting requirement. Every water system must identify and submit service line material data.
That distinction matters. Scope, completeness, and data quality differ substantially between surveys and inventories, and those differences directly affect how federal LSL replacement funding gets distributed. LCCR inventory data is now the best available national- and state-level information on lead service lines, which is why EPA appropriately shifted to using it for funding allocations.
Using the latest inventory data, EPA now estimates roughly 4 million LSLs remain nationwide, down from 9.2 million under prior survey-based estimates. This reduction is driven largely by methodological adjustments made to address reporting gaps. Notably, if a state failed to report inventory data for a system, EPA assumed that all service lines in that system are non-lead. While this approach likely underestimates remaining LSLs in some cases, it also creates a strong incentive for more complete reporting.
EPA also revised how it estimated non-lead lines for systems that reported total service connections but did not provide non-lead counts, subtracting reported lead, galvanized, and unknown lines from the total. Together, these changes corrected unreliable counts for several states, including Florida, and produced a more defensible national distribution of funds.
National Trends
Coverage and sample size
The updated 7th DWINSA survey contained responses from 3,520 water systems. By comparison, the SDWIS Service Line Inventory Dashboard reports data from 56,982 systems. The shift from a voluntary survey to a mandatory inventory increased coverage nearly sixteenfold, producing a more comprehensive picture of LSL prevalence nationwide. Data accuracy should continue to improve as systems update inventories annually.
Reported materials
Initial inventories paint a cautiously encouraging picture: more than half of reporting systems indicate they have no lead service lines. That said, some systems remain behind in completing inventories, and reporting practices vary.
Nationwide:
71,706,524 service lines are reported as non-lead. Reporting non-lead lines remains optional until November 1, 2027.
23,554,381 service lines are reported as having an unknown material type.
1,973,940 lead service lines and 969,625 galvanized service lines require replacement, consistent with EPA’s estimate of roughly 3 million reported LSLs or galvanized lines requiring replacement, plus approximately 1 million unknown lines predicted to be lead.
System-level reporting patterns
Among reporting systems:
1,113 systems (2%) did not account for any service line materials, despite indicating that required fields were complete.
36,992 systems (65%) reported all service lines as non-lead.
20 systems (<1%) reported all service lines as lead.
152 systems (<1%) reported all lines as galvanized requiring replacement.
3,060 systems (5%) reported all lines as unknown.
15,645 systems (27%) reported a mix of material categories.
State Trends
The statistics below reflect reported service lines at the state level; some systems in each state have yet to submit inventories.
Inventory data reveals substantial variation across states:
Oregon, Alaska, California, Colorado, and Washington have identified materials for more than 90% of reported service lines.
Maryland, New Hampshire, New Mexico, and Pennsylvania have identified material type for fewer than 50% of reported lines.
Most states report relatively low lead burdens. Thirty-eight states (76%) report that fewer than 5% of their reported service lines are lead or galvanized materials requiring replacement. Twelve states—Illinois, Wisconsin, Rhode Island, Michigan, Ohio, Indiana, Kansas, New York, Nebraska, Minnesota, Iowa, and New Jersey—report that between 5 and 13% of their lines are lead or galvanized requiring replacement. EPIC classifies states with roughly 8-12% lead or galvanized requiring replacement as heavily lead-burdened.
Two states stand out for exceptionally high shares of unknown materials: Maryland (99% unknown) and New Hampshire (98% unknown). Both received only the minimum 1% funding allotment EPA is required to give each state in this round, with Maryland’s share reduced by 0.4% compared to prior years.
Figure 1: Reported Service Line Material Distribution
What Changed? Comparing Survey and Inventory Data
To assess what was gained by shifting from surveys to inventories, we compared the proportion of reported service line material categories across datasets. Unlike the 7th DWINSA, the LCCR dataset does not include an “unreported” category. To enable an apples-to-apples comparison, we normalized the updated 7th DWINSA proportions to the total number of lines, excludingunreported lines.
A clear benefit of the LCRR inventory ismore current and granular data on lead service lines. Progress is evident in the number of lines previously classified as “unknown” that are now identified as non-lead. In some states, full replacement of lead service lines appears attainable within a reasonable timeframe. At the same time, Figure 2 reflects only reported service lines. Some systems have yet to report, and unreported lines may ultimately prove to be lead or galvanized requiring replacement. The extent of underreporting - and the likelihood that unreported lines include lead - varies significantly by state.
Figure 2: Lead Service Line Material Proportions: Inventories vs. Updated 7th DWINSA
In Figure 2, we grouped state-level changes into three patterns:
Good news (14 states):
Unknown materials shifted to non-lead. Hawaii, Minnesota, Tennessee, and Montana showed reductions greater than 50%. We expect more states to follow this pattern as inventories improve.
Concerning (3 states):
Increases in unknowns paired with decreases in non-lead reporting. Wyoming, New Hampshire, and Maryland fall into this group. This pattern likely reflects newly identified systems with unreported lines (i.e., more systems with incomplete inventories) or more conservative classifications under heightened scrutiny.
Mixed bag (33 states):
Reductions in lead and galvanized lines alongside increases in both non-lead and unknown categories. Some have hypothesized that utilities may be classifying lead or galvanized lines as “unknown” to avoid creating customer concerns or triggering notification requirements, particularly in cases where the actual line material has not been physically verified and could turn out to be non-lead. Overall:
Thirty-seven states increased reporting of non-lead lines
Eighteen increased reporting of unknown materials.
Marginal increases in galvanized lines requiring replacement occurred in Nebraska (3.32%), Alabama (2.39%), Michigan (1.45%), Kansas (1.44%), Mississippi (1.10%), and Missouri (1.06%).
Rhode Island saw a modest but notable increase in reported lead lines (3.25%).
North Dakota, Ohio, Wisconsin, Arkansas, and Illinois reduced reported lead lines by more than 10%.
The Path Forward: Turning Data Gaps Into Action
Reporting compliance remains uneven. Roughly one-third of water systems in New Mexico have not submitted inventory data, and New York still has more than 935,000 unreported service lines. Questions persist around resource constraints, accountability mechanisms, and enforcement.
Data coverage is also inconsistent. Statewide totals for service lines are not uniformly available, and “unknown” classifications remain prevalent: roughly one-third of states report that more than 30% of their service lines have unknown material status
These gaps have real consequences. Inventory quality directly affects community prioritization, utility replacement timelines, and state and federal funding allocations, especially now that EPA ties the Infrastructure Investment and Jobs Act (IIJA) lead service line replacement funding to inventory data.
EPIC is focused on clarifying these gaps and identifying practical solutions to support nationwide LSL replacement within the next decade. Key questions remain: How should the probability of lead among unknown or unreported lines be estimated? What constitutes “complete enough” inventory data? How does improved data quality translate into faster and fairer funding deployment? EPIC will continue to analyze these issues to ensure resources are used as effectively as possible to get the lead out.
