Is well water safe to drink? What the national data show, and how to know for your well

The short answer: well water can be perfectly safe to drink, and many wells are. But nobody can tell you that yours is without a lab result, because no government agency tests it for you. EPA's private wells page says the quality and safety of water from private domestic wells are not regulated by the federal government under the Safe Drinking Water Act, nor by most state governments, and that "private well owners are responsible for delivering safe drinking water to their households." CDC puts it even more plainly: "Tap water from private wells is not regulated, treated, or monitored by officials."

So "is well water safe to drink?" has two honest answers. In general: usually, but not always, and the problems that matter most are mostly invisible. For your well: you find out with a test from a state-certified lab, repeated on a schedule. Below: what the national data show (including what the often-quoted "23 percent" does and doesn't count), who needs to be most careful, how a well compares with city water, and what boiling and filters can and can't fix.

Why the answer is "it depends"

Two statements from official sources sit side by side, and both are true.

The reassuring one: the U.S. Geological Survey, citing EPA, writes that "most ground water in the United States is generally considered of good quality and safe to drink" (USGS Circular 1332). Wisconsin's health department opens its treatment fact sheet with "While most private wells in Wisconsin provide safe drinking water, some may contain substances that can affect our health" (Wisconsin DHS P-03494).

The cautionary one: the same USGS report found contaminants above health benchmarks in more than one in five sampled wells, and it adds that contaminant levels "can vary over short distances and with depth within aquifers," which points "to the importance of water-quality testing of individual wells." In practice that means your neighbor's clean result says little about yours. Years of drinking the water without trouble don't settle it either, because, as the next sections show, the contaminants tied most closely to health are mostly ones you can't taste or smell.

Behind both sits a gap in who checks. CDC's drinking water safety page says "The regulations that protect public drinking water systems do not apply to privately owned wells" and "you are responsible for making sure your private well water is safe to drink." EPA says about 15 percent of the U.S. population, over 43 million people, rely on private wells. For all of them, the only routine safety check is the one they order themselves.

What the national data show, and what "23 percent" really means

A large national study of untreated domestic-well water is the USGS National Water-Quality Assessment study of wells sampled from 1991 to 2004, published as Scientific Investigations Report 2008-5227 with a plain-language summary in Circular 1332. Its headline number is easy to quote; here is what is inside it.

The figure comes from Table 10 of the report: 314 of 1,389 wells (22.6 percent) had at least one contaminant above a human-health benchmark. Those 1,389 wells, from 25 principal aquifers in 45 states, are the subset in which major ions, trace elements, nutrients, radon and organic compounds were all measured. The Circular rounds it to "More than one in five (23 percent)."

Four details change how you should read it:

• Bacteria are not in it. Fecal indicator bacteria were measured in too few wells to include. The report says that given how often bacteria were found and "the benchmark of no bacteria detections," the real share of wells with any contaminant above a benchmark "likely would have been higher."
• It depends on the radon yardstick. The 22.6 percent counts radon against 4,000 pCi/L, the higher of two levels EPA proposed in 1999 and never finalized. Against the lower proposed 300 pCi/L, 64.9 percent of the same wells had radon alone above that level. There is still no federal radon-in-water standard; our radon explainer covers what states use instead.
• Most of it is natural. In the report's words, "19.2 percent of wells contained at least one naturally occurring contaminant (radon, a trace element, or fluoride) at a concentration greater than its benchmark." Rock and soil, not the homeowner, put them there.
• Low levels are almost universal. 93.9 percent of wells had at least one contaminant above one-tenth of a benchmark, mostly because radon is counted here against one-tenth of the lower proposed level (30 pCi/L), and radon was detected in 97.4 percent of wells. That is not a health finding by itself, since a tenth of a limit is below the limit, but it shows how rarely well water is "pure."

Mixtures are the open question. Only about 4 percent of wells had two or more contaminants above their benchmarks, but about 73 percent had mixtures of contaminants each above one-tenth of its benchmark. The authors write that "the long-term cumulative effects of low concentrations of multiple contaminants on human health currently are unknown."

Keep the authors' own caveats in mind. The wells "are not a statistically representative sample of all domestic wells in the United States." Samples were taken before any home treatment, so they describe the source water, not what comes out of a filtered tap. And the sampling ended in 2004. Use the study to see which problems are common, not to estimate your own odds.

Which contaminants turn up, and where they come from

USGS found eight inorganic contaminants above health benchmarks in 1 percent or more of domestic wells. The table that lists them carries a one-line verdict: "Except for nitrate, these contaminants are derived primarily from natural sources."

That split matters for what you do next. Contamination from people (septic systems, fertilizer, fuel) tends to have a nearby cause you can sometimes find and fix. Natural contaminants such as arsenic and radon don't. USGS notes that they "may be present ... in relatively undeveloped settings that otherwise may not be perceived as vulnerable to contamination," and that "traditional well-head protection approaches to preventing contamination generally are not effective" against them. A well in the woods with no farm in sight can still have arsenic. See the arsenic explainer for where it is most common.

Man-made organic compounds (pesticides, solvents, fuel components) were detected in 60 percent of the wells, the Circular says, but above benchmarks in less than 1 percent. Detection is common; exceeding a health benchmark is rare.

Bacteria are the other big category. In the report's Table 6, total coliform bacteria were detected in 133 of 397 wells tested (33.5 percent) and E. coli in 30 of 378 (7.9 percent). Total coliforms are an indicator group; E. coli, as the Circular puts it, "typically is not harmful but is an indicator of fecal contamination and, therefore, the possible presence of pathogens." The coliform bacteria explainer walks through what a positive result means.

The report also gives a sense of the illness side: though few in number (averaging fewer than 20 a year), 20 to 40 percent of the drinking-water disease outbreaks reported to CDC from 1999 to 2002 were attributed to domestic well water, and the authors add that gastrointestinal illness linked to well water "is likely underreported."

Finally, much of what owners notice is not a health issue. About half (48 percent) of the sampled wells had at least one property outside EPA's aesthetic ranges, such as pH, dissolved solids, iron or manganese, the kind of thing that stains fixtures or tastes off.

Can well water make you sick?

Yes. EPA says germs from sewage and animal waste can cause gastrointestinal illness and infections, high nitrate can make infants below six months seriously ill, and metals, organic chemicals and radionuclides can damage organs or raise cancer risk at high levels. Most give no warning you can taste, so only a lab test shows whether your well is the cause.

Timing differs. For babies, nitrate can act fast: EPA calls "blue baby syndrome" an "acute condition" that "can occur rapidly over a period of days." Most chemical effects are long-term; EPA's drinking water table lists them as effects "from long-term exposure above the MCL" unless it says otherwise. Those limits are written for public water systems and don't bind a private well. The table below is EPA's wording, contaminant group by group:

Sources: EPA, Potential Well Water Contaminants and Their Impacts (updated July 6, 2026); EPA, National Primary Drinking Water Regulations (updated August 31, 2026).

MDH lists the symptoms of disease-causing microorganisms in drinking water as "diarrhea, vomiting, cramps, nausea, headaches, fever, fatigue, and even death sometimes." Symptoms alone can't show where an illness came from, and a clean-tasting glass doesn't rule the well out. If someone in the house is ill and you suspect the water, Wisconsin's health department says: "Talk to your health care professional if you or a family member have symptoms that you believe are related to your drinking water." Then have the well tested for bacteria and nitrate at a certified lab; the coliform and E. coli pages cover what to do with a positive result.

What you can sense, and what you can't

Here is the uncomfortable part. The things you can notice are mostly the aesthetic ones, and the contaminants most tied to health are mostly the ones you can't. The Minnesota Department of Health says that "you cannot taste, see, or smell most contaminants" in groundwater (MDH, Bacteria in well water).

Even the clues people trust can mislead. Penn State notes that corrosive water can produce "small leaks, blue-green stains, or metallic tastes" in copper plumbing but that these "are not reliable indicators of the risk of lead in water." Clear, cold, good-tasting water is not evidence of anything.

The reverse is also true: a bad smell or a stain is often a nuisance rather than a hazard, though some changes (a sewage or fuel smell, or water that changes after heavy rain) are reasons to stop drinking until tested. Our smell, taste and stains guide sorts those out symptom by symptom.

Who should be most careful

Agencies single out the same groups again and again. The wording below is theirs; where a decision depends on your health or pregnancy, they send you to a doctor or the health department, and so do we.

That federal wording is written for utility reports; a well owner gets no such notice. For babies, the nitrate rules are the sharpest, and the nitrate explainer covers them in detail.

Well water vs. city water: what regulation does and doesn't cover

People often ask whether well water is safer or less safe than city water. None of the federal and state pages we read for this guide says one is safer than the other. What they describe is a difference in who checks, and that difference is concrete.

EPA says the Safe Drinking Water Act exists "to protect public health by regulating the nation's public drinking water supply," and that "SDWA does not regulate private wells which serve fewer than 25 individuals" (EPA SDWA overview). A public water system, in EPA's definition, serves at least 15 service connections or an average of at least 25 people for at least 60 days a year (EPA, public water systems). EPA counts over 148,000 of them, supplying 90 percent of Americans.

Regulation means somebody is required to look. An unregulated well is not automatically worse; it is unexamined until you examine it.

How to know if your well is safe: five steps

1. Get the baseline test. CDC and EPA both say to test at least once a year for total coliform bacteria, nitrates, total dissolved solids and pH, and CDC says to "Use a state-certified laboratory to test your water." Our lab finder lists certified labs from official state lists. The how-to-test guide covers sampling, which is where many results go wrong.
2. Add the tests your area and household call for. The yearly four do not include arsenic, uranium, radon, manganese or lead. Minnesota recommends arsenic, lead and manganese "At least once." CDC says to "Contact your health department to find out what other germs or chemicals to test for based on where you live." Our which-tests checklist builds a list from EPA and CDC triggers (farmland, old plumbing, gas stations, a baby on the way).
3. Read the report against the right benchmark. Put the numbers into the results checker, which places each one next to EPA's limits. Check the units and the reporting limit before you compare; the how-to-test guide explains both, and the lead explainer covers first-draw lead samples.
4. Retest to confirm before big decisions, and after any fix. EPA says that if a contaminant exceeds health standards, contact your public health department for specific steps and have the well retested to confirm it. Minnesota says to "Test your water again after taking action," and, if you treat your water, to test it after treatment.
5. Keep testing. Yearly for the baseline, sooner when something changes. EPA's list of reasons to test right away includes known problems with groundwater in your area, conditions near your well that "have changed significantly," repairs to any part of the well system, and a change in odor, color or taste. CDC adds a pregnancy and "A child starts living with you."

EPA notes that your local health department "may provide private well testing for free." Our testing cost guide lists real prices from state and county labs.

Does boiling make well water safe?

Only for germs, and sometimes it makes other things worse.

CDC's drinking water advisories page says "Boiling kills germs in the tap water that could make you sick": bring water to a full rolling boil for 1 minute, or 3 minutes at elevations above 6,500 feet. The same page says "Boiling water containing harmful chemicals or toxins will not make the water safe to use."

With nitrate, boiling backfires. Minnesota: "Do not try to boil nitrate out of the water. Boiling will make nitrate more concentrated." Water evaporates and the nitrate stays behind. MDH's bacteria page carries the same warning for a common combination: "If nitrate or arsenic were also detected in your well water, boiling your water may not be the safest option."

A worked example. Your report says total coliform present, E. coli absent, nitrate 4.2 mg/L as N. The nitrate is below the 10 mg/L standard, but it was detected, so per MDH boiling "may not be the safest option" for drinking water until the bacteria problem is fixed. Bottled water or another safe source fits better while you disinfect the well and retest. That is row 5 of the decision table below. If nitrate had not been detected, MDH lists boiled or bottled water for cooking and drinking as acceptable in the meantime.

Filters: point-of-use, whole-house, and what they can and can't do

Treatment can make a problem well drinkable, but only treatment matched to a test result, and only if it is maintained. Minnesota's health department states it bluntly: "No single treatment unit can remove all contaminants in water" (MDH, Home water treatment). It also says to first try removing the source of contamination or replacing the contaminated supply.

There are two basic layouts. MDH: "Point-of-use (POU) units treat water at one faucet or one location," while "POE units treat all of the water in your home." Wisconsin's health department explains when each fits: POU devices are "best when the water contains substances that pose a risk when it is used for drinking and preparing foods," and POE systems are best when substances "pose a risk when it is used for bathing, showering, and running the dishwasher and washing machine."

Wisconsin's fact sheet also lists which device type is certified to reduce contaminants commonly tested for in that state, and which NSF/ANSI standard to look for. Six of its twelve rows:

Source: Wisconsin DHS, Treatment Devices for Private Well Contaminants (P-03494, 11/2024). Wisconsin also notes that softener salt "can increase the sodium level in water."

A certification only counts for the contaminant it names. MDH says to select a unit certified by NSF, UL or WQA "to remove the contaminant(s) you are concerned about, if a certification is available," and warns that "These organizations do not certify treatment units for all contaminants." Bacteria are a different job: MDH's steps for a positive coliform result are to disinfect the well with a chlorine solution and retest; a filter is not among them.

Maintenance is part of the safety. MDH: units that are not maintained "will lose their effectiveness over time," and "In some cases, unmaintained units can make water quality worse and can make you sick." Wisconsin says to test the treated water after a plumbed-in device is installed.

Can you drink well water without a filter? Many people do, and nothing in the CDC, EPA or state guidance we read says a well needs treatment when its results are below the benchmarks. What that guidance asks for is the testing, and a clean result only covers what was tested. A filter bought without a test may target the wrong thing: Wisconsin's table lists no single device for every contaminant, and a softener, for example, is listed only for manganese.

Decision table: your situation and what to do now

The table is a summary of the agencies' own guidance, not medical advice. When a result is close to a limit, or when a baby, a pregnancy or an illness is involved, call your local health department; it knows what turns up in wells in your area.

Questions people ask