Somewhere in a kitchen drawer or an unopened email folder sits a document listing everything your water utility found in your tap water last year. Most people never open it. The ones who do often give up inside a minute.
That is a failure of translation, not a failure of attention. When researchers at Virginia Tech scored a national sample of 30 of these reports, they found them written at an 11th to 14th grade reading level, far above the 6th to 7th grade level recommended for public health communication. A companion study scored the same reports against the CDC’s Clear Communication Index. The average came in at 50 percent, no report reached the 90 percent qualifying threshold, and only 63 percent said outright whether the water was safe to drink.
The document has a formal name: the Consumer Confidence Report. Your utility may print it as the annual drinking water quality report. Federal rule requires it and fixes what goes in it and where. Once you know the structure, ten minutes is enough.
This article walks through that structure, and through the part that matters most: where the report’s coverage stops.
Who has to send one, who has to get one, and by when
The obligation falls on one kind of utility. EPA requires a report from every community water system, meaning a public water system serving the same people year-round. Your city utility is one. So is a suburban water district or the system in a mobile home park.
EPA’s consumer page states the deadline in thirteen words: “Reports must be sent by your water supplier each year by July 1.” The regulation says the same thing at 40 CFR 141.152(b), and the report delivered in July covers the previous calendar year. You are always reading history.
The delivery rule is stricter than most people assume. Under 40 CFR 141.155, a community water system has to deliver a copy directly to each customer. Mailing paper works, as does mailing or emailing a direct link to the report online. Ask for paper and the system has to provide it. The system then certifies to its state regulator within 10 days that it distributed the report, and keeps a copy for at least three years, which is why most utilities can hand you an old one on request.
Renters are the gap the rule tries to close. If you do not pay a water bill, you are a consumer rather than a customer, and the system has to make a good faith effort to reach you anyway. The regulation’s examples run from posting the report online to advertising it or holding a public meeting. In practice that often amounts to a web page you have to go looking for, which is why most renters have never seen one.
A change is coming. Congress told EPA to revise the rule in the America’s Water Infrastructure Act of 2018, and EPA announced the final revision in May 2024. A system serving 10,000 or more people will distribute the report twice a year instead of once, with the second delivery due by December 31, starting in 2027. The original rule dates to August 19, 1998, so this is the first real overhaul in a quarter century.
Then there are the people the rule never covered. EPA counts more than 148,000 public water systems, but only community systems send these reports. A school, factory or highway rest stop running its own well is a different class of system and sends you nothing. If you draw from a private well, nobody federal checks your water: EPA’s overview of the Safe Drinking Water Act says the Act does not regulate private wells serving fewer than 25 people. A separate article here covers that split.
Six terms that look interchangeable and are not
Here is where most readers lose the thread. The report leans on six pieces of vocabulary that all sound like versions of “the limit.” They are not versions of each other. They sit in different legal categories, and the category tells you how much weight a number carries.
The pair that sets numbers: MCL and MCLG
A maximum contaminant level, or MCL, is the enforceable one. EPA’s national primary drinking water regulations define it as the highest level of a contaminant allowed in drinking water, and the page states directly that MCLs are enforceable standards. It is defined in regulation with consequences attached: exceed it and your utility is in violation, has to tell you, and has to fix it.
A maximum contaminant level goal, or MCLG, is the health target. EPA’s definition, which your report is required to reproduce word for word, is “The level of a contaminant in drinking water below which there is no known or expected risk to health.” Nothing enforces it. EPA calls MCLGs non-enforceable public health goals.
The gap between the two is the most instructive thing in the whole document. Arsenic has an MCLG of zero and an MCL of 0.010 mg/L. Benzene has an MCLG of zero and an MCL of 0.005 mg/L. EPA sets the goal where the health science points, then sets the enforceable limit where treatment and laboratory measurement can reach. That trade-off is made in public, and it has a cost: the enforceable number is not the number a toxicologist would pick in a world without budgets. A zero in the MCLG column does not mean your utility failed. It means the goal is zero and the law requires something achievable.
The pair with no enforcement behind it: secondary standards and health advisories
National secondary drinking water standards cover 15 contaminants that affect taste, color, odor, staining and equipment rather than health. EPA describes them as non-mandatory water quality standards. Iron at 0.3 mg/L, total dissolved solids at 500 mg/L, pH between 6.5 and 8.5: these are the numbers behind the orange ring in a sink and the metallic taste. EPA does not enforce them, with one narrow exception, a public notice requirement when fluoride exceeds 2.0 mg/L.
A health advisory is the other item with no enforcement, and it does a different job. EPA issues advisories to provide what the agency calls a non-regulatory human health-based level of protection, typically for a contaminant that has no national regulation yet. States and utilities use them to make decisions before a rule exists. A number labeled “health advisory” in your report is information, not a ceiling.
Both belong to the government advisory category rather than the regulatory one: real federal numbers with nothing behind them but EPA’s judgment. Your utility can sit above either without breaking any law.
The pair that describes a required action: treatment technique and action level
Some contaminants cannot be measured at the tap reliably enough to set a numeric ceiling. For those, EPA requires a process instead of a number. A treatment technique, in EPA’s wording, is a required process intended to reduce the level of a contaminant in drinking water. Giardia, Legionella and enteric viruses are handled this way, and the MCL column for them is empty by design rather than oversight. A treatment technique is regulation: skipping the process is a violation just as surely as exceeding an MCL.
An action level is a trigger inside one of those processes, and it is the term readers most often mistake for an MCL. Lead and copper are regulated by treatment technique, and the action level is the tap-sample concentration that forces the system to do more. EPA’s current lead action level is 15 parts per billion. If more than 10 percent of tap samples exceed it, the system must act, through corrosion control, public education and replacing the lead service lines it controls. EPA’s table notes the level drops to 0.010 mg/L on November 1, 2027 under the Lead and Copper Rule Improvements. Exceeding an action level is not automatically a violation. It starts a clock. A separate article covers lead in detail.
The two tables, and the order to read them
Find the violations first.
The rule requires any violation to be clearly flagged in the data, with an explanation a normal person can follow, covering how long it lasted, the health effects involved, and what the system did about it. If there were no violations, the report says so near the top.
Then go to the detected contaminants table. Every regulated contaminant the utility found has to appear there with its MCL, its MCLG and the level detected, expressed in the same units. Contaminants tested for and not found do not appear at all. That is why the table is usually short, and why its shortness is not suspicious.
Read it in three columns: what was found, what the enforceable limit is, and how far apart those two numbers are. The distance carries the information. A contaminant at 10 percent of its MCL is a different situation from one at 90 percent, and both are legal.
Range, average, and the footnote where the news hides
A single figure in a results column is almost never a single measurement.
For contaminants monitored annually or less often, the rule requires the report to give the highest level detected at any sampling point along with the range of detected levels. For contaminants judged on a running annual average, such as disinfection byproducts, the report has to give the highest average at any monitoring location plus the range of individual sample results. Unregulated contaminants get an average and a range.
So a table showing 2.1 with a range of 0.4 to 7.8 is telling you something an average alone would bury: somewhere in the system, at some point in the year, a sample came back at 7.8. Compliance may rest on the average. Your exposure on a given Tuesday does not.
Footnotes are where the awkward material goes. A system operating under a variance or an exemption, which is permission not to meet a standard under certain conditions, has to explain the reason, the issue date and its compliance status. A monitoring violation, meaning the utility failed to test when it was supposed to, usually appears here rather than in bold type. So does the note that a result came from testing done in an earlier year, because some contaminants are monitored on a multi-year cycle. If anything in the report is going to change how you feel about your water, it is more likely to be in small type at the bottom than in the headline.
Parts per million, billion and trillion
Units are the quietest source of confusion in the document, because they make very different quantities look alike on the page.
EPA’s regulation table reports in milligrams per liter, mg/L, and states that mg/L is equivalent to parts per million. One part per billion is a thousandth of that, which is why the lead action level of 0.015 mg/L is the same quantity as the 15 parts per billion EPA writes on its lead page. One part per trillion is a thousandth again, and EPA’s PFAS regulation page writes 4.0 parts per trillion as the equivalent of 4.0 nanograms per liter.
A time comparison makes the spacing physical. One part per million is about one second out of 11.6 days. One part per billion is about one second out of 32 years. One part per trillion is about one second out of 32,000 years.
That is not a reason to shrug at small numbers. A few contaminants matter at parts per trillion, which is precisely why EPA regulates them there. It is a reason to check the unit before you react to the digits, because the same quantity looks frightening in ppt and harmless in ppm.
The part the report cannot tell you: your own building
This is the section worth keeping if you forget everything else.
The Consumer Confidence Report describes a water system. It does not describe your house.
Your utility samples inside its own system, and the results describe water it produced and distributed. Between the main in the street and the glass in your hand sits the service line onto your property, your building’s plumbing, the solder in its joints, the brass in its fixtures, and in an apartment building a riser and storage tank the utility has never inspected. EPA’s own page on lead says the most common sources of lead in drinking water are lead pipes, faucets and fixtures, which is to say the parts that no annual report covers.
Time in the pipe matters too. A 2023 review in Science of the Total Environment calls premise plumbing, the pipework inside a building, the hardest part of the chain to manage, with inconsistent conditions that can shift the water’s microbial community. A 2024 field study of 204 building plumbing systems in the International Journal of Hygiene and Environmental Health found elevated cell counts at taps and pipe ends where water had been standing, and a Legionella contamination rate 22.3 percent higher in the systems it called microbiologically unstable.
Neither study is about your kitchen. Both make the same structural point: the last stretch of pipe is its own problem, and the annual report has nothing to say about it.
Practical consequences follow. If your building is older, a clean lead row in the report does not clear your tap. If you want to know what comes out of your faucet, the only route is testing your own water, which another article in this library covers. And if water has been sitting in the pipes overnight or through a vacation, run the cold tap before you drink or cook.
“No violations” is not “nothing detected”
Two sentences in these reports get misread constantly, and they get misread in opposite directions.
“No violations” means the utility met every enforceable standard. It does not mean the water contained nothing. Treated water from a surface source will contain measurable disinfection byproducts, because disinfection is the thing keeping pathogens out and the byproducts are what that costs. A detection below an MCL is the ordinary condition of treated drinking water, not a warning sign.
“Not detected” also means less than it sounds like. It means the concentration was below the level the laboratory method could reliably report. It does not mean zero. Detection limits move as methods improve, which is one reason a contaminant can appear in this year’s table after years of absence without anything changing in the water itself.
For scale, EPA states on its Safe Drinking Water Act page that more than 92 percent of the population served by community water systems receives water meeting all health-based standards all of the time. That is a high number and not a perfect one, and your report tells you which side of it you are on.
A few findings should send you to a person. If the report lists a violation with a health effects statement, call the phone number the report is required to carry and ask what the system is doing and whether the advice applies to your household. If anyone at home is pregnant, immunocompromised, or an infant drinking formula mixed with tap water, raise the specific contaminant with a clinician instead of deciding alone. Where Cryptosporidium monitoring applies, the report carries an explanation written for people with weakened immune systems, and that paragraph has a real audience. Urgent problems also do not wait for July: a boil water notice arrives on its own schedule and overrides anything in the annual report.
A ten-minute reading routine
Work through it in this order the next time one lands.
- Find the date range. A July report covers the previous calendar year, so you are reading a year that has already ended.
- Check the violations statement first. If there were none, the rest of the document is information rather than news.
- Identify the source. Surface water, groundwater, or water purchased from another system. It explains most of what follows.
- Scan the detected contaminants table for any result in the same neighborhood as its MCL, not only for results above it.
- Compare the range against the single reported figure. The spread tells you whether the system runs steady or swings.
- Read every footnote. It takes two minutes, and variances, exemptions and monitoring failures live there.
- Check the unit before you react to the number. ppm, ppb and ppt differ by factors of a thousand.
- Ask what the report cannot answer. Your service line, your building’s plumbing and your own tap sit outside its scope.
One more step costs nothing. Save the report, and next July put the two side by side. A single year’s table tells you whether a system was compliant. Two or three years of tables tell you whether anything is drifting, which is the question this document answers best and the one almost nobody asks it.
Keep exploring: water filter and water testing.
Sources
- U.S. Environmental Protection Agency. “Consumer Confidence Reports (CCR): Information for Consumers.” epa.gov
- U.S. Environmental Protection Agency. “National Primary Drinking Water Regulations.” epa.gov
- U.S. Environmental Protection Agency. “Secondary Drinking Water Standards: Guidance for Nuisance Chemicals.” epa.gov
- U.S. Environmental Protection Agency. “EPA Non-Regulatory Health-Based Drinking Water Levels.” epa.gov
- Electronic Code of Federal Regulations. “40 CFR Part 141, Subpart O: Consumer Confidence Reports.” ecfr.gov
- Electronic Code of Federal Regulations. “40 CFR 141.153: Content of the reports.” ecfr.gov
- Electronic Code of Federal Regulations. “40 CFR 141.155: Report delivery and recordkeeping.” ecfr.gov
- U.S. Environmental Protection Agency. “Consumer Confidence Report Rule Revisions.” epa.gov
- U.S. Environmental Protection Agency. “Consumer Confidence Report Rule and Rule History for Water Systems.” epa.gov
- U.S. Environmental Protection Agency. “Information about Public Water Systems.” epa.gov
- U.S. Environmental Protection Agency. “Safe Drinking Water Act (SDWA).” epa.gov
- U.S. Environmental Protection Agency. “Overview of the Safe Drinking Water Act.” epa.gov
- U.S. Environmental Protection Agency. “Basic Information about Lead in Drinking Water.” epa.gov
- U.S. Environmental Protection Agency. “Per- and Polyfluoroalkyl Substances (PFAS) Drinking Water Regulation.” epa.gov
- PubMed. “Roy S, et al. “An evaluation of the readability of drinking water quality reports: a national assessment.” Journal of Water and Health, 2015. PubMed. DOI: 10.2166/wh.2015.194.” doi.org
- PubMed. “Phetxumphou K, et al. “Assessing clarity of message communication for mandated USEPA drinking water quality reports.” Journal of Water and Health, 2016. PubMed. DOI: 10.2166/wh.2015.134.” doi.org
- PubMed. “Abkar L, et al. “Microbial ecology of drinking water from source to tap.” Science of the Total Environment, 2023. PubMed. DOI: 10.1016/j.scitotenv.2023.168077.” doi.org
- PubMed. “Genuardi MD, et al. “Statistical analysis of parameters affecting Legionella and total cell growth in premise plumbing systems within buildings.” International Journal of Hygiene and Environmental Health, 2024. PubMed. DOI: 10.1016/j.ijheh.2024.114456.” doi.org