Lead in water is measured through several different lenses: infrastructure projections, tap-sample compliance results, groundwater chemistry, school testing programs, and childhood blood-lead surveillance. The figures below keep those populations, units, dates, and definitions separate so that a service-line estimate is not mistaken for a household exposure rate.
Key Lead in Water Statistics
The most broadly useful figures are:
- In the 2023 EPA report, 9,223,745 U.S. service lines were projected to contain lead, covering the United States, territories, Puerto Rico, and the District of Columbia.
- In the same 2023 EPA report, 99,949,560 total service lines were projected nationally, including projected material categories.
- The 2023 EPA report projected about 9.2% of the national service-line total had lead content, calculated from its two projected counts rather than reported as a separate prevalence estimate.
- EPA projected 2,800,839 standalone galvanized service lines that had never been downstream of lead in the 2023 report.
- EPA projected 87,929,975 service lines with no-lead content in the 2023 report.
- EPA estimated $50 billion to $80 billion in lead-service-line replacement costs, in 2021 dollars, for the United States and territories covered by the report.
- The EPA report estimated $5,328 as the low replacement cost per lead service line and $9,015 as the high estimate, in 2021 dollars.
- The 2023 EPA update projected 9,031,938 lead-content service lines in the United States.
- In 2024, 147,252 active public water systems were included in EPA’s compliance-report universe.
- In 2024, 45,980 active public water systems had at least one reported drinking-water-standard violation in the United States.
- In 2024, 10,769 active public water systems had at least one reported health-based drinking-water-standard violation.
- The WHO’s 2022 fact sheet gives 10 µg/L as a provisional guideline value for lead in drinking water.
- WHO states that drinking-water lead concentrations are generally below 5 µg/L, while concentrations above 100 µg/L have been measured where lead service connections or fittings are present.
- About 44 million people relied on self-supplied groundwater for drinking water in a 2019 U.S. study summary.
- About 33% of more than 8,300 untreated U.S. groundwater samples had modeled lead-solubility potential above 15 µg/L in samples collected from 2000 through 2016.
- EPA-supported programs allocated $180.5 million from 2019 through 2024 to states, territories, and tribes for reducing children’s lead exposure in drinking water.
- The same programs reported 8.8 million children benefiting from 2019 through 2024.
- CDC receives about 3 million childhood blood-lead test results each year, a surveillance volume rather than a national prevalence estimate.
Contents
- National lead service-line statistics and exposure scale
- Lead service-line data by state and replacement cost
- Lead in tap water measurements and regulatory benchmarks
- Private-well and groundwater lead statistics
- Lead testing in schools and child-care facilities
- Blood-lead surveillance and water-related evidence
National Lead Service-Line Statistics and Exposure Scale
EPA’s 2023 Drinking Water Infrastructure Needs Survey and Assessment is a national infrastructure projection, not a direct census of contaminated taps. It includes known, unknown, and unreported lines across the United States, territories, Puerto Rico, and the District of Columbia.
| Measure | EPA projection | Period and scope |
|---|---|---|
| Lead-content service lines | 9,223,745 | 2023 report; 2021-dollar baseline; U.S. and covered territories |
| Total service lines | 99,949,560 | 2023 report; same geography |
| Standalone galvanized lines never downstream of lead | 2,800,839 | 2023 report; same geography |
| No-lead-content lines | 87,929,975 | 2023 report; same geography |
Source: Drinking Water Infrastructure Needs Survey and Assessment 7th Report to Congress.
The ratio of the projected lead-content count to the projected total is about 9.2%, but that is a calculation from EPA’s two reported projections, not a separate EPA prevalence estimate. The report’s categories also mean that an infrastructure classification does not by itself prove that every connected household has contaminated water.
The 2023 update reported a somewhat different national table: 9,031,938 projected lead-content service lines out of 100,182,373 projected total service lines. Because the update uses its own known, unknown, and unreported line information, its figures should be read as an updated projection rather than automatically merged with the earlier table.
Source: 2023 Update to the 7th Drinking Water Infrastructure Needs Survey and Assessment.
State Lead Service-Line Data and Replacement Cost
State projections show how unevenly the estimated infrastructure burden is distributed. The percentages below are shares of the national projected total in EPA’s 2023 report, not direct inventories.
| State | Projected lead service lines | Share of national projection |
|---|---|---|
| Florida | 1,159,300 | 12.62% |
| Illinois | 1,043,294 | 11.35% |
| Michigan | 301,790 | 3.28% |
| Missouri | 202,112 | 2.20% |
| Arkansas | 171,771 | 1.87% |
| Connecticut | 146,574 | 1.60% |
| Colorado | 111,907 | 1.22% |
| Alabama | 91,544 | 1.00% |
| Hawaii | 9,589 | 0.10% |
Source: Drinking Water Infrastructure Needs Survey and Assessment 7th Report to Congress.
EPA estimated the national replacement cost at approximately $50 billion to $80 billion in 2021 dollars. Its lower per-line estimate was $5,328, based on the 25th percentile of reported replacement-cost data, while its higher estimate was $9,015, based on the 75th percentile.
These are planning estimates based on projected line counts and replacement-cost data. They are not invoices for every property, and actual project costs can vary by location and project conditions.
Lead in Tap Water Measurements and Regulatory Benchmarks
Lead concentrations and compliance rules use different measurement systems. WHO’s 2022 chemical fact sheet gives a provisional guideline value of 10 µg/L, while the U.S. EPA lead action level is 0.015 mg/L, or 15 µg/L.
WHO says concentrations in drinking water are generally below 5 µg/L, but concentrations above 100 µg/L have been measured where lead service connections or fittings are present. WHO describes its value as provisional and based on treatment performance and analytical achievability, not as a health-based threshold.
Sources: Guidelines for Drinking-water Quality: Lead Chemical Fact Sheet; What the 90th percentile means in a public water system report.
| U.S. compliance measure | Definition | Geography and period |
|---|---|---|
| Lead action level | 0.015 mg/L, equivalent to 15 µg/L | U.S. public water systems; EPA explanation |
| System exceedance | More than 10% of tap samples above 0.015 mg/L | U.S. public water systems; EPA explanation |
| Five-sample 90th percentile | Average of the fourth- and fifth-highest results | Systems permitted to collect five samples |
| Fewer-than-five-sample rule | Highest result is used | U.S. public water systems allowed fewer samples |
The action level triggers regulatory actions; it does not mean concentrations below it are risk-free. For systems with five samples, EPA’s 90th-percentile calculation uses the fourth- and fifth-highest results, while systems allowed fewer than five samples use the highest result.
EPA recorded 147,252 active public water systems in calendar year 2024. Of those, 101,272 had no reported drinking-water-standard violations, 45,980 had at least one reported drinking-water-standard violation, and 10,769 had at least one reported health-based violation; the latter categories include contaminants and treatment-technique failures beyond lead.
EPA also reported about 3,902 systems with health-based violations tied to the 2024 lead-service-line-inventory requirement, or about 2.7% of active systems. That is an inventory-requirement violation count, not a count of systems with measured lead contamination.
Source: Providing Safe Drinking Water in America: National Public Water Systems Compliance Report.
Private-Well and Groundwater Lead Statistics
Private wells and self-supplied groundwater are a separate exposure context from regulated public water systems. A 2019 study summary estimated that about 44 million people in the United States relied on self-supplied groundwater for drinking water, but that population estimate does not mean all of those people are exposed to lead.
USGS assessed groundwater samples collected from 2000 through 2016. Its occurrence dataset included 13,324 U.S. groundwater sites, while the modeled dataset included 8,313 sites with complete required chemistry.
Source: Assessing the lead solubility potential of untreated groundwater of the United States.
- About 33% of more than 8,300 untreated groundwater samples had modeled lead-solubility potential above 15 µg/L.
- Five percent of untreated groundwater samples had modeled lead-solubility potential above 300 µg/L.
- Less than 1% of more than 8,300 untreated groundwater samples had measured dissolved lead above 15 µg/L.
The modeled values describe how much lead could dissolve under the sampled chemistry if lead-bearing plumbing were present; they are not measured tap-water concentrations. The measured well-water result is also not a complete household exposure measure because plumbing can add lead after the wellhead.
USGS identified Maryland, New Jersey, Delaware, Tennessee, and Arkansas among the states with the greatest share of wells having modeled lead-solubility potential at or above 300 µg/L, without giving one common percentage for that state group in the summary.
Source: Domestic groundwater wells in the eastern and southeastern U.S. at risk of lead contamination.
Lead Testing in Schools and Child-Care Facilities
EPA-supported programs measure implementation activity rather than national health outcomes. From 2019 through 2024, programs allocated $180.5 million to states, territories, and tribes and reported that 8.8 million children benefited.
| Program output | Reported total | Period and scope |
|---|---|---|
| Child-care facilities tested | 20,435 | 2019–2024; U.S. states and territories |
| Child-care facilities completing remediation | 1,380 | 2019–2024; program-reported actions |
| Schools tested | 12,810 | 2019–2024; U.S. states and territories |
| Schools completing remediation | 2,506 | 2019–2024; program-reported actions |
| Guides, factsheets, and tools | 27 | 2020–2024; EPA program |
Source: By the Numbers – Reducing Lead Exposure to Children in Drinking Water.
Across the two facility categories, EPA reporting counted 33,245 facilities tested, a sum of the reported school and child-care totals. Facilities may differ in size and sampling intensity, and completed remediation actions do not quantify residual lead or prove a measured reduction in exposure.
Blood-Lead Surveillance and Water-Related Evidence
CDC receives about 3 million childhood blood-lead test results each year through national surveillance. CDC lowered the childhood blood-lead reference value from 5 to 3.5 µg/dL in October 2021, with classifications using 2022 data; the reference value prioritizes action and is not a toxicity threshold or water-specific attribution measure.
Source: Childhood Blood Lead Surveillance: National Data.
CDC’s surveillance is not population-based and cannot produce national prevalence estimates. Blood lead also cannot be attributed to drinking water from these data alone, so surveillance volume and the reference-value change should not be presented as measurements of water contamination.
An EPA Office of Inspector General review identified 498 water systems with lead action-level exceedances from tap samples reported between January 2021 and March 2023. Those systems served about 2.3 million people, according to the review, but the figure is the population served by identified systems in EPA tracking data, not a national exposure estimate.
USGS also reported radionuclide findings from filtered untreated groundwater samples in 19 principal aquifers in the contiguous United States. Lead-210 exceeded its 1.0 pCi/L health-concern benchmark in 3.7% of samples from 1,263 public-supply wells, while polonium-210 exceeded its 0.7 pCi/L benchmark in 1.5% of those samples.
Among groundwater samples with radon-222 above 4,000 pCi/L, 72% had lead-210 at or above 1.0 pCi/L. These are radionuclide measurements distinct from ordinary dissolved lead measured in µg/L, and the conditional association is not a population-wide risk estimate.
Source: Assessing the lead solubility potential of untreated groundwater of the United States.