Is Tap Water Safe to Drink?

In fiscal year 2023, 92.4% of people served by US community water systems got their water from a system with no reported violation of EPA's health-based standards all year (EPA's Report on the Environment). That is a real, reassuring baseline. It is not the same as proof that every glass was risk-free.

Community water systems are the ones that serve the same homes year round, which is why this figure speaks to your kitchen tap more directly than the broader count that also folds in campgrounds and gas stations. More than 320 million people, roughly 95% of the country, get their water this way. "Health-based standards" is the plain-English shorthand for the EPA rules meant to protect health directly: a limit on how much of a contaminant is allowed (an MCL), a limit on disinfectant levels (an MRDL), and required treatment steps (a treatment technique). A system with no reported violation of any of those, across a full year, is the situation 92.4% of served people were in. That share has improved over time; EPA’s comparison point was 78.8% in fiscal 1993.

The figure describes systems and years, not individual glasses. "No reported violation" means a system stayed within those standards as measured and reported, not that the water was flawless at every address and every hour.

The same EPA source also shows where the year fell short, and it is worth seeing plainly rather than glossing over. The largest categories were surface-water treatment violations (systems serving 14.4 million people, 4.5%) and disinfection-byproduct violations (3.7 million, 1.2%); coliform-rule violations reached 2.7 million people (0.85%), and arsenic and nitrate violations were smaller still (312,800 and 175,300 people respectively). Those are real shortfalls, but they are concentrated in particular systems and years, not a sign that the baseline is hollow.

Two cautions keep that counterweight honest. When a system has a violation, EPA counts its entire service population even though only part of it may actually receive non-compliant water, so these exposure numbers can overstate how many people drank affected water; EPA itself notes there is no way to know how many, if any, people were truly drinking water in violation. And because a few systems are very large, a single one can move the national figure by two or three points in either direction. Separately, some systems fall short on monitoring or reporting (late, incomplete, or missing data) rather than on a health standard itself. That is tracked apart from confirmed health-based violations, but it can leave the picture less complete than a single percentage suggests.

So the honest reading of the baseline is steady, with gaps that matter. That clean compliance record answers the legal question. The health question is separate, and it comes next.

Not necessarily. EPA sets two numbers for many contaminants: an enforceable legal limit (the MCL) and a health goal (the MCLG). When they differ, the gap is usually about what treatment can achieve, measure, and afford, not a loophole or proof regulators stopped caring. So "within the limit" is not the same as "100% safe."

A report that shows no violations, like the national baseline above, is telling you a system stayed within its enforceable limits. That is real, and it matters. It is also a legal statement more than a complete health one. In plain terms, the health goal is the level EPA considers to carry no known risk, with a margin built in, and the enforceable limit is the line a system can actually be held to. Often the two are the same number. When they are not, it is usually because the health goal sits below what current treatment can reliably remove, or even measure, at a cost systems can bear. The law directs EPA to set the enforceable limit as close to the health goal as is feasible. The goal points to where the health science leads; the limit reflects what is achievable right now.

Arsenic is the cleanest illustration. EPA's enforceable limit is 10 micrograms per liter, the same figure the World Health Organization uses, but EPA's health goal for arsenic is zero (EPA National Primary Drinking Water Regulations; WHO arsenic fact sheet). That looks like a contradiction and is not one. WHO labels its own 10 a provisional value precisely because lower levels are hard to reach and hard to measure reliably, and EPA's limit reflects the same practical ceiling. The zero goal says no amount of arsenic is treated as free of risk; the enforceable 10 says that is the lowest point treatment can dependably hold today.

PFAS, the so-called forever chemicals, show the same split in a live and shifting form. In 2024 EPA finalized its first national PFAS drinking-water standards, six in all, and for two of them, PFOA and PFOS, set the enforceable limit at 4 parts per trillion against a health goal of zero (PFAS National Primary Drinking Water Regulation, Federal Register). As with arsenic, the zero is not a claim that any trace causes harm; it reflects EPA finding no level it can identify as risk-free, with the enforceable number set at what testing and treatment can currently manage.

The status here is in motion, so it is worth dating. As of June 3, 2026, all six of those 2024 standards remain final and on the books. EPA proposed two changes in May 2026: one proposed rescission rule that would roll back four of the six standards, and one proposed PFOA/PFOS compliance extension rule that would let certain systems request more time to meet the PFOA and PFOS limits. Both are proposals, not final rules, with comments due July 20, 2026. Unless and until something is finalized, the 2024 limits all still stand, including the standards proposed for rollback and the original compliance timing. (What these chemicals actually do at different exposure levels is a separate question, covered further down.)

This also explains how two official sources can publish different numbers for the same chemical without either being useless. A regulator setting an enforceable limit is answering what systems can be held to under the drinking-water law. A health agency may be answering a different exposure question, using a different time frame, endpoint, or safety assumption. Different questions can produce different numbers. Knowing which source is answering which question is its own skill.

No, not fully. A city or system meeting federal standards is necessary, but it does not settle what is true at your particular faucet. Compliance is measured across a whole system, while some of the biggest risks, lead especially, are added after the water leaves treatment, inside your service line and household plumbing.

It helps to picture the water's path in two stretches. The first is the part your water system controls: the source, the treatment plant, and the large mains running through town. The second is the stretch closest to you, the service line into your building and the pipes, solder, and fixtures inside it. Federal standards are mostly tested and enforced on that first stretch. The second is largely on your side of the meter, and it is where the most common household problem appears.

Lead is where this gap matters most, and it is regulated unlike a contaminant such as arsenic for one main reason: it usually does not come from the source water at all. It enters later, when water meets lead service lines, lead solder, older brass fixtures, or galvanized pipe, and corrosion lifts some of it into what comes out of the faucet (EPA, Lead Service Lines). Because the lead is picked up so close to the tap, EPA cannot manage it with a single number measured at the treatment plant. So rather than a conventional enforceable limit, lead is handled through required treatment steps, mainly corrosion control, alongside a long-term program to locate and replace lead service lines. EPA's health goal for lead is zero. No level is treated as safe, and the concern is sharpest for young children and during pregnancy.

That structure is also why a system can be meeting the lead rules while a particular home still has elevated lead. Compliance is judged from a 90th-percentile figure. A system samples a set of higher-risk homes, and the rule looks at the level that 90 percent of those samples fall below; if that figure crosses a set threshold, called the action level, the system must strengthen corrosion control and take further steps. By design, then, up to a tenth of sampled homes can sit above that threshold while the system stays within the rules, and your home may share nothing with the homes that were sampled. A citywide pass is real information. It is not a reading from your kitchen sink.

These rules are mid-transition, which is worth knowing when you read anything local. EPA's Lead and Copper Rule Improvements are final and in effect, and they lower the lead action level to 10 parts per billion under the revised framework, down from the older 15 (EPA LCRI overview; Federal Register final rule). But most of the major obligations, including the schedule for replacing lead service lines, formally begin on November 1, 2027. Until then, your local water report may still describe lead using the legacy 15 ppb framework, so a current report and the newer standard can both be accurate while looking different. The rule is also being challenged in court by water-industry groups, which is a reason to keep an eye on its details, not a reason to discount the lead concern or to treat the rule as optional. It remains in effect.

In practice the takeaway is narrow and usable. If you live in a home built before 1986, when lead plumbing and solder were still common, if you do not know what your service line is made of, if you are mixing infant formula, if you are pregnant, or if you are uneasy about one specific fixture, then citywide compliance is useful background rather than a personal answer. Closing that gap is a local and household question: what your specific area and building tend to carry, and, where it is warranted, a direct test of the water coming out of your tap. For now, the important point is simpler: citywide compliance is useful context, but your home may still need its own answer.

Not usually. For most healthy adults on a regulated public system with no active violations, advisories, or high-risk plumbing, tap water is generally safe to drink and unlikely to be a major added health risk. Risk concentrates around specific contaminants and specific people: infants, pregnant women, the immunocompromised, private wells, and lead-prone homes.

The useful version of this question is not "is tap water bad" but "under what conditions, and for whom." Most of the contaminants people worry about follow a pattern: at the low levels typical of a compliant system, they are usually not a major added risk for most healthy adults, while at an exceedance, during an advisory, or with years of higher exposure they matter much more. That is the legal-versus-health and system-versus-home split from above, now applied to your body. The contaminants below are the ones worth understanding, each with its real level of concern stated plainly.

Disinfection byproducts

When a water system adds chlorine or a similar disinfectant to kill germs, a small share of it reacts with natural organic matter already in the water and forms new compounds. Two regulated groups, trihalomethanes (TTHM) and haloacetic acids (HAA5), are the most closely watched. The concern with them is mainly long-term: exposure above the limits over many years has been associated with a possible increase in cancer risk and with liver, kidney, or nervous-system effects (EPA National Primary Drinking Water Regulations).

Disinfection is one of the main reasons waterborne disease is uncommon in regulated US systems, and the rules try to keep byproducts low without weakening protection against germs. So this is a long-term balance a system manages, not a reason to fear chlorinated water or to want disinfection stopped. The limits exist to keep the byproducts low while the disinfection keeps doing its job.

PFAS

PFAS, a large family of long-lasting industrial chemicals, are a real long-term concern, and also a place where a single detection is easy to misread. EPA's current understanding links certain PFAS exposures to increased risk of some cancers (including kidney, testicular, and prostate), reduced immune and vaccine response (the most sensitive effect, and one seen in children), reproductive and developmental effects including lower birth weight, thyroid and hormone disruption, and higher cholesterol (EPA, Our Current Understanding of the Human Health and Environmental Risks of PFAS).

That list comes with two honest qualifiers. EPA has not set a cancer-risk concentration in water for PFOA or PFOS, because those cancer analyses are still ongoing. And the health goal of zero does not mean any detectable amount causes harm; it means EPA has not identified a level it can call risk-free, which is a different statement. As covered earlier, all six of the 2024 PFAS standards remain final as of June 3, 2026, with the May 2026 changes still only proposals. If your water report or a local source flags PFAS, that is a reason to learn your actual numbers and act on them, not a reason to assume harm has already occurred.

Nitrate

Nitrate breaks the usual pattern, because its most serious risk is acute rather than slow. EPA's limit is 10 milligrams per liter, measured as nitrogen (EPA Chemical Contaminant Rules). The group at real risk is narrow but important: infants under about six months, mainly when formula is mixed with high-nitrate water. In those infants, nitrate can interfere with the blood's ability to carry oxygen, a condition called methemoglobinemia, or "blue-baby syndrome," which can become serious or fatal without prompt care. Older children and adults are far less sensitive.

Nitrate is also a know-your-local-water contaminant rather than a national worry. It concentrates where fertilizer runoff, septic systems, and animal waste reach groundwater, which puts agricultural areas and private wells at the front of the line.

One warning is important enough to state plainly: do not boil high-nitrate water to make it safer for infant formula. Boiling does not remove nitrate, and it actually concentrates it as water evaporates. If high nitrate is a possibility, use a different, known-safe water source for formula and contact your water provider or local health department.

Microbial contamination and boil-water advisories

Microbial contamination, from bacteria, viruses, or parasites, is one of the clearest short-term drinking-water risks, and it is the main reason advisories exist. A boil-water advisory is often precautionary: it can be issued when water is, or simply could be, contaminated after a main break, a loss of pressure, a treatment problem, or flooding, not only when contamination has been confirmed. It does not mean everyone has already been drinking unsafe water.

It helps to know the three kinds of advisory, because each calls for a different response (CDC, Drinking Water Advisories: An Overview). A boil-water advisory means boiling will make the water safe to use. A do-not-drink advisory points to something boiling will not fix, often a chemical, so bottled water is used for drinking and cooking until it lifts. A do-not-use advisory means avoiding the water altogether, including for washing, until it is cleared.

When boiling is the instruction, the guidance is a rolling boil for one minute, or three minutes above 6,500 feet of elevation. Boiling reliably kills viruses, bacteria, and parasites. It does not remove chemical contaminants, which is exactly why the do-not-drink and do-not-use categories exist.

Lead

Lead belongs on this list too, but it is the clearest case of household-specific rather than system-wide risk, so the detailed explanation belongs in the earlier section on whether your city’s compliance reaches your tap. The short version: it matters most for young children and during pregnancy, and because it usually enters from home plumbing and service lines, the real question is about your home, not the national average.

Two quick notes round out the picture. Fluoride is not a simple good-or-bad contaminant: US public-health agencies still recommend community fluoridation around 0.7 mg/L for cavity prevention, while newer neurodevelopmental reviews raise concern mainly at higher total exposures, especially above about 1.5 mg/L; that trade-off deserves its own full guide.

And because filters come up constantly with these contaminants, it is worth saying clearly: a filter is not a substitute for knowing the contaminant, but when testing, a CCR (your annual local water report), a lead service line, or an advisory points to a specific risk, a certified filter can be the right interim tool. Matching a filter to a particular problem is a separate decision.

The national baseline does not apply to you. EPA regulates public water systems, but it does not regulate private household wells, so no agency is monitoring your water or reporting violations on your behalf. Testing it, maintaining it, and deciding whether it needs treatment are the owner's responsibility.

A private well can be excellent water, and many are. But "excellent" and "verified" are different things, and a well is not covered by the monitoring, treatment requirements, and annual public reporting that sit behind a community water system (EPA, Private Drinking Water Wells). The national baseline from the top of this page simply does not describe it.

What a well is exposed to depends on local geology and what sits nearby. The common concerns are bacteria, nitrate, and arsenic, and the odds rise with agricultural runoff, septic systems, flooding, and nearby land use. None of these announce themselves at the tap, which is the same lesson as the section above: you find them by testing the water, because taste alone cannot reveal them.

At a minimum, most guidance points to testing a private well at least once a year for bacteria and nitrate, with added tests for arsenic or other local concerns depending on your area's geology and land use. If you are mixing infant formula, the nitrate point from earlier applies directly and is worth acting on rather than assuming. The specifics of what to test for and how often depend on your well, your area, and nearby land use.

Not categorically. Bottled water is regulated differently, not automatically better or worse. FDA standards generally track EPA contaminant standards, but testing, source protection, public reporting, cost, plastic exposure, and emergency usefulness differ. So the honest answer depends on your situation rather than on the fact that it came in a bottle.

The difference starts with who is in charge. EPA regulates public tap water under the Safe Drinking Water Act, while the FDA regulates bottled water as a packaged food. By law, the FDA's bottled-water standards must be at least as protective of public health as the EPA's, and they match EPA's limits for dozens of substances; on lead, the bottled standard is actually stricter than the tap limit (FDA, FDA Regulates the Safety of Bottled Water).

In practice, though, oversight of bottlers is looser in several ways. A 2009 GAO review found bottled-water protections were often less stringent than comparable tap-water protections: bottlers test less frequently than large municipal systems, there is no equivalent mandate to protect the source water, and bottled water sold only within one state can fall outside federal oversight entirely (GAO). The source is sometimes the same water. By NRDC's estimate, a quarter or more of bottled water starts as municipal tap water, sometimes treated further and sometimes not, though the exact share is debated (NRDC).

So bottled water is not a routine upgrade just because it comes sealed in plastic, and it carries a higher cost and environmental footprint. Where it earns its place is specific and short-term: during an active boil-water or do-not-drink advisory, when a household has a known lead problem, or for an immunocompromised household during an advisory. Plastic exposure, including microplastics, is a real and active question, but it belongs with the fuller comparison rather than here. For this page, the key point is enough: bottled water can be useful in specific situations, but it is not automatically a better everyday answer.

Usually, sources disagree because they are answering different questions. Regulators like EPA set enforceable floors, the limits a system can be held to. Health agencies may focus on health effects, exposure context, or screening values. Advocacy groups may focus on lower health-based targets or worst-case concerns. Neither is lying; the trick is knowing which question a given number is answering.

A simple ranking helps, less as a pecking order than as a map of what each source is for.

Start with EPA's enforceable standards, and with your local water report as their local form. These tell you whether your system is actually meeting the limits it can be held to. That is the legal floor, and it is the most concrete thing you have.

Next come the health goals: EPA's own MCLGs, and the World Health Organization's health-based guidelines where they differ, as with arsenic. These describe where the health science points, separate from what is enforceable today. They are why "within the limit" and "no known risk" can be different statements.

Then there are the federal health agencies, the CDC and ATSDR, which publish health context and screening values, often for emerging concerns like PFAS. These sources help explain health effects, exposure context, vulnerable groups, and screening values, which is different from saying what a water system can be held to under the drinking-water law.

Advocacy sources like the Environmental Working Group belong on the map too, as useful context rather than the final word. EWG's database often looks far more alarming than an EPA table, and the reason is not dishonesty: it measures against its own health-based targets, which sit well below enforceable limits, and it leans toward flagging concern early. Local news works similarly, surfacing real problems while compressing the nuance. Treat both as a prompt to look closer, not as the verdict itself.

One test holds the hierarchy together. Before trusting a framing, ask whether both an EPA scientist and a mainstream toxicologist would call it fair. A claim only one camp would endorse, in either direction, is a signal to slow down. Different sources can each be useful without being interchangeable, and the skill is matching the source to the question you actually have. When you are ready to move from “is tap water generally safe?” to “what should I check in my own water?”, the Home Water Quality hub gives you the step-by-step path: source, local report, testing, interpretation, and treatment decisions.