Your water looks fine. It is clear in the glass, it does not smell like anything, and it has tasted the same for the eleven years you have lived in the house.
None of that is information. Taste, smell and appearance are not reliable guides to whether water is safe. Lead is tasteless. Nitrate is colorless. Arsenic has no odor. The contaminants most worth knowing about are the ones your senses were never built to catch.
Testing is the step almost everyone skips, and it is the one that tells you whether any of the rest of the advice applies to you. Buying a filter before you have a result is buying a solution to a problem you have not identified yet.
One piece of context decides everything downstream. EPA sets and enforces standards for public water systems under the Safe Drinking Water Act, while private household wells sit under no federal regulation at all. Which side of that line you are on changes your whole testing job, and this library’s article on municipal versus private well water takes the distinction apart in full.
Who should test, and when
If you are on a private well, EPA and CDC land on the same baseline: once a year, every year. EPA’s annual short list is total coliform bacteria, nitrates, total dissolved solids and pH. CDC gives the same four, plus whatever else is common in your local groundwater, which your health department can name for you.
Both agencies say to test more often when the household includes small children, an older adult, or someone pregnant or nursing.
Then there are the events that should trigger a test regardless of the calendar. Sort them by what changed.
Something changed in the water. The taste, color or smell shifted. Fixtures or laundry started staining, which points toward iron, copper or manganese. Someone in the house has had repeated gastrointestinal illness, which points toward coliform bacteria.
Something changed around the well. Flooding, land disturbance, nearby construction, or new industrial activity. A known water problem reported in your area. Intensive agriculture close by, which argues for nitrate, nitrite, pesticides and coliform bacteria. Gas drilling nearby, which argues for chloride, sodium, barium and strontium.
Something changed in the system. Any part of the well replaced or repaired. Plumbing work inside the house, including a new water heater or repiping.
Two more belong on the list though the agencies do not frame them as triggers. The first is moving in. You have inherited a water supply whose history you do not know, and a baseline test in month one beats any seller’s assurance. The second is a pregnancy or a new baby, because formula is made with tap water and a baby drinks a large volume relative to body weight.
Public water customers have a smaller job. Your utility tests the system and reports annually, and this library’s article on interpreting water quality reports covers that document. What the utility’s testing cannot tell you is what happens in the pipes between the street and your tap, which is exactly where lead gets in. If your house is old, or you do not know what your service line is made of, your own tap is the only place that answer lives.
For scale: EPA’s private wells page cites a USGS study of roughly 2,100 private wells in which about one in five contained at least one contaminant above a human health benchmark. That is not a scare number. It is a base rate, and it means a well that has never been tested is a coin flip you have not looked at.
What a certified laboratory does that a strip cannot
Certification is a real mechanism, not a sticker, and it is worth understanding before you spend money.
EPA’s Office of Water Technical Support Center runs the Drinking Water Laboratory Certification Program alongside EPA regions and the states. To hold certification, a laboratory has to be certified by EPA or by its state, use EPA-approved analytical methods, pass on-site audits periodically, and pass proficiency testing at least once a year for every method and every analyte it wants certified.
Proficiency testing is the part that protects you. A provider sends the lab a sample whose true concentration the lab does not know. The lab runs it and reports a number. If that number falls outside the acceptable window, the lab loses certification for that substance. It is a graded exam, annually, one subject at a time.
State certification is a government program written into federal rules. That is a different category from a private certification mark on a box, and a very different category from “lab tested” printed on packaging, which has no defined meaning at all.
Now the strips, where the evidence is specific and unkind. In 2021, Kriss and colleagues at Virginia Tech and Northeastern University published a study in Environmental Science and Technology evaluating 16 brands of at-home lead-in-water test kits. Twelve of the sixteen were not suitable for drinking water analysis, with detection limits between 5 and 20 milligrams per liter. Lead in drinking water matters at the microgram-per-liter scale, which is a thousand times smaller. Those kits were blind at the level that counts.
Binary strip kits did detect dissolved lead at drinking-water-relevant concentrations. They failed on particulate lead, which is lead carried as tiny solid flakes rather than dissolved. Adding a household acid such as lemon juice or vinegar dissolved enough of those particles to push soluble lead into detectable range. That is a useful laboratory finding, not a consumer instruction.
So here is the honest division of labor. A strip is reasonable for a rough read on pH, hardness or chlorine, the properties you can partly sense anyway. A strip is not a basis for deciding whether your child’s drinking water is safe.
One more thing costs people results: the laboratory’s own instructions beat the kit’s packaging every time. Sample bottles for some analytes come pre-dosed with acid or sodium thiosulfate. Bacteria samples need a sterile container and a short window to reach the lab. Collect the sample wrong and the lab will still hand you a number, and the number will be about your technique.
First draw or flushed changes the answer
This is the technical point that most changes a result, and homeowners are almost never told about it.
Water sitting still in a pipe is in contact with that pipe. Metal migrates into it over hours. Water that just came through at full flow spent seconds in contact with the same pipe. Same house, same tap, same morning, two very different numbers.
EPA’s 3Ts guidance for schools and child care facilities lays out a two-step protocol, and the logic carries straight over to a house.
Step one is the first draw. Water must sit unused for at least 8 hours and no more than 18. You collect 250 milliliters the instant the tap opens, before anyone has run water anywhere in the building. Do not take the aerator off first. Do not flush first. Do not close a shutoff valve. The sample is meant to capture what is sitting in that fixture, so disturbing any of it defeats the purpose.
Step two is the flush sample. EPA uses it when the first draw comes back high. You run the tap at a steady flow for 30 seconds, then collect another 250 milliliters. Because that water came from farther back in the plumbing, comparing the two results points to where the lead is entering: the fixture itself, or the pipe behind it.
EPA uses a small 250 milliliter sample for two reasons worth knowing. A small volume represents a short length of pipe, which is what lets you localize the source. And it is close to the amount a person drinks at once.
Two practical consequences follow.
First, a result means nothing without the protocol attached to it. If anyone hands you a number and cannot tell you whether it was a first draw or a flushed sample, and how long the water sat, that number is not comparable to anything, including a standard.
Second, flushing is not a guaranteed fix. A 2015 study by Pieper and colleagues in Environmental Science and Technology profiled 15 private water systems and found three different patterns. In some, lead showed up only in the first draw, and 15 to 30 seconds of flushing brought it below 15 micrograms per liter. In others, lead arrived as erratic spikes of particles, and flushing did not reliably help. In a third pattern, lead stayed detectable continuously, because it was coming from corroding components inside the well itself. Advice to “just run the tap” is right for one of those three and wrong for the other two.
And test every tap you drink from, not one. Redmon and colleagues collected 1,266 first-draw samples across 86 North Carolina child care centers for a 2020 study in Environmental Research. About 77 percent had detectable lead at or above 0.1 micrograms per liter, and concentrations varied widely between water points inside the same building. The kitchen tap and the bathroom tap are not the same test.
Where lead comes from, what the current federal rule requires, and what to do about a high result each belong to other articles here. The point in this one is narrower: the sample you take determines the answer you get.
How to read what comes back
A laboratory report is a table, and three ideas make it readable.
Detection limit. Every analytical method has a floor below which it cannot reliably tell a substance apart from zero. That floor is a property of the method and the instrument, not of your water. The North Carolina study could see lead at 0.1 micrograms per liter. Most of the home kits in the Virginia Tech study could not see below 5 milligrams per liter. Both of those are detection limits, and they sit fifty thousand times apart.
Non-detect. A non-detect, usually printed as ND or as a less-than sign followed by a number, does not mean zero. It means the substance was not found above the detection limit of the method used. A result of less than 1 microgram per liter and a result of less than 20 micrograms per liter say completely different things, and the second is nearly useless if what you care about would show up at 5. Always find the detection limit sitting next to the non-detect.
Units. This is where readers lose the thread entirely, so it gets its own section.
Parts per million, billion and trillion are not near neighbors
They read like a tidy sequence. They are a cliff.
- One part per million is one in a million. EPA states on its drinking water regulations table that milligrams per liter are equivalent to parts per million.
- One part per billion is one in a billion, a thousand times smaller. It is the same quantity as a microgram per liter.
- One part per trillion is one in a trillion, another thousand times down. EPA’s PFAS drinking water limits are written in parts per trillion, which EPA also expresses as nanograms per liter.
Arithmetic you can check yourself makes the gap concrete. Round a drop of water to one twentieth of a milliliter. That drop in about 13 gallons of water is roughly one part per million. The same drop in a 13,000 gallon backyard swimming pool is about one part per billion. That drop spread across twenty Olympic-size pools is about one part per trillion.
This matters on a real report because different substances are regulated at wildly different scales. EPA publishes its drinking water table in milligrams per liter by default, so nitrate appears as parts per million. Lead is discussed in parts per billion. PFAS compounds are regulated in parts per trillion. Glance at “0.004” on one line and “4.0” on another and you may conclude the first is smaller, when the units can make it exactly the reverse.
So the rule is simple, and it is worth making a habit. Never compare two numbers on a water report until you have checked that the units match. Convert everything into one unit first. One part per million equals 1,000 parts per billion equals 1,000,000 parts per trillion.
Free tests, utility tests, and who pays
Three different things get called “a water test,” and they are not interchangeable.
The free in-home test from a company that sells treatment equipment. Be clear-eyed about what it is. It is a sales qualification step. The technician typically checks hardness, total dissolved solids, chlorine and pH, which happen to be exactly the properties a softener or a filter visibly improves. A hardness reading is a real hardness reading, so the measurement is not necessarily wrong. But the panel was chosen to produce a recommendation, and it is not a certified laboratory analysis of the substances that affect health. If the visit ends with a quote, the quote was the point.
Your local or state health department. EPA and CDC both send people here first, and it is badly underused. Many health departments test private wells for coliform bacteria, nitrate and sometimes more, free or cheap, and they will tell you which contaminants show up in your area’s groundwater. That local knowledge is the part you cannot buy anywhere else.
A state-certified laboratory. This is the real one, and you pay for it. EPA tells well owners to send samples only to laboratories carrying drinking water certification, and it maintains contact information for every state’s certification program. CDC gives the Safe Drinking Water Hotline, 800-426-4791, as a route to a certified lab near you. EPA is blunt about its own role: “EPA is unable to test residential or commercial drinking water upon request.” Nobody federal is coming to your kitchen.
Costs vary by panel. EPA puts a lead test on your own tap at roughly $20 to $100. A broad well panel runs higher. You can hold the cost down by testing annually for the basic four and adding the area-specific items your health department names, rather than buying a sprawling panel every year.
Getting it done, start to finish
A plan you can run inside a week.
- Work out what you are on. Private well, public system, or a shared or community well. If you do not know, your county health department does.
- Call the health department first. Ask two questions: what do you test for free, and what contaminants turn up in this area’s groundwater? Write the answers down.
- Pick a state-certified lab from your state program’s list. Tell them about the house: age, known plumbing materials, well or city water.
- Ask the lab for sampling instructions before they ship bottles, and read them. Ask specifically whether a lead sample should be first draw, and what stagnation window they want.
- Sample every tap you drink from and cook from, not just the kitchen.
- Set a calendar reminder twelve months out. An annual test only works if it is annual.
- When the report arrives, check units before conclusions, and look for the detection limit beside every non-detect.
When to stop reading and make a call
If a result comes back above a health standard, contact your public health department rather than a retailer. EPA’s advice is to get next steps from them and have the well retested to confirm the finding before you act on it.
Call the same day, not at your convenience, if someone in the house has been repeatedly ill with gastrointestinal symptoms and you are on a well, if the water changes suddenly in taste, color or smell, or if your well has been flooded. Treat a flooded well as contaminated until a test says otherwise.
If your area is under a boil water advisory, follow the advisory and do not wait on your own test result. And if someone is sick, that is a question for a clinician, not a laboratory. A water test tells you about the water. It does not diagnose anybody.
The part worth keeping
The order is the whole lesson: find out what is in your water, then decide what to do about it. Reverse that order and you are buying equipment to solve a guess. What to do about a result, including which treatment technology matches which contaminant, is the subject of its own article here, and it only becomes answerable once you have a number.
In practice the work is one phone call to a health department, one state-certified laboratory, sampling instructions read before you touch the tap, and a reminder set for next year. For most households that is under two hundred dollars, often a good deal less.
When the report arrives, check the units twice. The factor of a thousand between parts per million and parts per billion is the most common way a homeowner misreads a result that was perfectly clear.
Sources
- U.S. Environmental Protection Agency. “Private Drinking Water Wells.” epa.gov
- U.S. Environmental Protection Agency. “Protect Your Home's Water.” epa.gov
- U.S. Environmental Protection Agency. “Certification of Laboratories for Drinking Water.” epa.gov
- U.S. Environmental Protection Agency. “Learn about Laboratory Certification for Drinking Water.” epa.gov
- U.S. Environmental Protection Agency. “Contact Information for Certification Programs and Certified Laboratories for Drinking Water.” epa.gov
- U.S. Environmental Protection Agency. “3Ts for Reducing Lead in Drinking Water.” epa.gov
- U.S. Environmental Protection Agency. “3Ts Module 5: 2-Step Sampling Procedures (PDF).” epa.gov
- U.S. Environmental Protection Agency. “3Ts Module 4: Understanding the Sampling Procedures (PDF).” epa.gov
- U.S. Environmental Protection Agency. “National Primary Drinking Water Regulations.” epa.gov
- U.S. Environmental Protection Agency. “Per- and Polyfluoroalkyl Substances (PFAS) and the Safe Drinking Water Act.” epa.gov
- U.S. Environmental Protection Agency. “Basic Information about Lead in Drinking Water.” epa.gov
- Centers for Disease Control and Prevention. “About Drinking Water.” cdc.gov
- Centers for Disease Control and Prevention. “Guidelines for Testing Well Water.” cdc.gov
- Centers for Disease Control and Prevention. “About Choosing Home Water Filters.” cdc.gov
- Environmental Science & Technology (PubMed). “Kriss R et al. "Challenges of Detecting Lead in Drinking Water Using at-Home Test Kits." Environmental Science & Technology, 2021. PubMed. DOI: 10.1021/acs.est.0c07614.” doi.org
- Environmental Science & Technology (PubMed). “Pieper KJ et al. "Profiling Private Water Systems to Identify Patterns of Waterborne Lead Exposure." Environmental Science & Technology, 2015. PubMed. DOI: 10.1021/acs.est.5b03174.” doi.org
- Environmental Research (PubMed). “Redmon JH et al. "Lead in drinking water at North Carolina childcare centers: Piloting a citizen science-based testing strategy." Environmental Research, 2020. PubMed. DOI: 10.1016/j.envres.2020.109126.” doi.org