Radon levels explained: what your pCi/L actually says
One unit, two thresholds, and a lot of misreading. Here’s what picocuries per liter measure, where EPA’s lines come from, and the honest action attached to every range your report could show.
Five minutes, no sign-up. Read this first, then save the result to a printable home record or work out what to ask a contractor. Not measured yet? Start with what a test kit settles.
The unit
A picocurie per liter (pCi/L) measures radioactive decays happening in a liter of air. Radon — an invisible, odorless gas produced by uranium decaying in soil and rock — exists outdoors everywhere at low background levels; the question a test answers is how much your house concentrates it. There is no absolute zero to chase, which is exactly why guidance works in thresholds rather than absolutes.
The two lines that matter
EPA recommends action at 4 pCi/L and suggests considering action between 2 and 4. The 4 pCi/L figure is a practical policy line — the level at which EPA judges reduction both warranted and reliably achievable — not a cliff where air becomes dangerous at 4.0 and safe at 3.9. Risk from radon is cumulative and roughly proportional to concentration and time; EPA identifies long-term exposure as the leading cause of lung cancer among people who never smoked. Lower is simply better, and 4 is where “fix it” becomes the standing advice.
Reading your result
One short-term test is a snapshot of one set of conditions, not a fixed property of the house. Readings move with the season, the weather, and how the house was operated during the measurement — which is why the bands below describe what to do next rather than what the house is.
| Result | What guidance says | Reasonable next step |
|---|---|---|
| Below 2.0 pCi/LNo action recommended | EPA recommends no action at this level. That is guidance about what to do, not a verdict on exposure — lower is still better, and EPA notes that some risk remains at any indoor concentration. | Retest every couple of years, and again after foundation, HVAC, basement, or air-sealing work. |
| 2.0 – 3.9 pCi/LConsider action | EPA suggests considering action in this band. It is a household judgement rather than a mandate, and reasonable people decide it differently depending on who lives there and how the lowest level is used. | A long-term detector over 90+ days tells you where in the band you actually sit before you commit to anything. |
| 4.0 pCi/L and aboveEPA recommends action | This is the action level. It is not an emergency and it does not require anyone to leave the house, but it is the point at which reducing the level is the recommended response. | Confirm the test conditions were valid, then get written quotes from certified contractors that include a post-mitigation verification test. |
Why one result is not the whole story: the measurement reflects the floor it was taken on, the conditions during the test, and the season. A winter reading under closed-building conditions is closer to your realistic upper end; a mild-weather reading is closer to your lower end. If your number sits near a band boundary, a second measurement is cheaper than the decision it informs.
Every range, with its honest next step
| Your result | What it means | Honest next step |
|---|---|---|
| Under 2 pCi/L | At or near typical background; no guidance calls for action. | File it. Retest in a couple of years or after major renovations. |
| 2 – 3.9 pCi/L | EPA’s consideration band — a judgment zone, weighed by basement use and household factors. | Long-term test for a truer average; act if the lowest level is heavily lived in. |
| 4 – 7.9 pCi/L | At or above the action level; the standing recommendation is to reduce it. | Confirm if unhurried, then quotes; skip confirmation inside a deal window. |
| 8 pCi/L and up | Well above the action level — same fix, more reason to schedule it. | Straight to licensed quotes; retesting rarely changes this decision. |
Reading a result when the test wasn’t textbook
Most people arrive at this page holding a number and a doubt: the windows were open, the kit sat on a sill, the test ran two days instead of seven. The useful question is not whether the test was perfect — almost none are — but which direction the imperfection pushes the reading, and what the reading still establishes anyway. Most of these conditions bias a result in a knowable direction, and knowing which way is what lets you hold a sensible conversation with a licensed professional instead of paying for a second test you may not need.
| If this happened during the test | Which way it could push the number | What the result still establishes |
|---|---|---|
| Device near a window or exterior door | The reading is likely to sit lower than a compliant placement in the same room would have produced, though by how much nobody can say. | A result above the action level from a diluted position is still a result above the action level, and arguably a stronger signal than the same figure from mid-room. |
| Tested in a kitchen or bathroom | The reading could sit above or below a living-area measurement on the same floor, and which way is not predictable from the description. | The device functioned and produced a number. What is in question is what the number represents, not whether it happened. |
| The device was moved | The reading sits somewhere between what the two positions would each have produced, weighted by time in each, which nobody recorded. | If a continuous monitor was used, the log may show the move and let someone read the periods separately, which recovers most of the value. |
| Tested during severe weather | A short test that included a significant pressure swing may sit higher than the same house would read over a settled week. | Weather does not manufacture soil gas. A high reading during a storm still indicates that the building draws soil gas when conditions favour it, which they will again. |
| A whole-house or attic fan ran | The reading probably sits below what closed conditions would have produced in the same period. | A reading above the action level taken while the house was being ventilated is a stronger indication than the same figure under closed conditions, not a weaker one. |
| The test was shorter than intended | The result could sit either side of what the full duration would have produced, and the shorter the exposure the wider that uncertainty. | A short exposure returning a high figure is still evidence that levels are present, and generally worth acting on to the extent of measuring properly. |
| Only an unoccupied level was tested | The occupied floors are likely to read lower, but by an amount that varies with the building and cannot be estimated from a description. | The basement measurement is the standard benchmark and is exactly what a professional or a buyer will ask for. It is not the wrong test. |
| Windows were opened | A result under a threshold might sit above it under closed conditions; that is the case where this matters most. | A high reading in a ventilated house is a strong indication rather than a suspect one. |
| The device was too low or too high | A floor-level reading is likely to sit above what a compliant height would have produced in the same room. | The measurement still demonstrates that soil gas is entering, which is the substantive finding. |
| Nobody recorded the conditions | Nothing about the physical measurement — the effect is entirely on what the result can be used to argue. | The reading is a real measurement of a real building and remains useful for the household own decision. |
| Comparing results from different devices | The gap between the two may reflect real variation rather than error, particularly across seasons. | Neither result is invalidated by the existence of the other, and the lower one is not automatically the correct one. |
| Treating one result as permanent | A house that read low years ago may read differently now, in either direction, with nothing visibly different about it. | The original measurement was valid for its period and remains a useful data point, particularly if its conditions were documented. |
What this table cannot do: it cannot tell you your true long-term average, quantify how far any condition shifted your specific reading, or substitute for a licensed measurement where one is required. It sorts what a result still supports from what it never established — the rest is a measurement question, and what a platform can and cannot determine remotely sets out where that line sits.
Why your number moves between tests
Radon entry rides on weather, soil moisture, frost, wind, and how the house was operated during the test — closed-house winter readings run higher than open-window June ones because the house holds more of what enters. Two honest tests can differ meaningfully. That’s not measurement failure; it’s why averages and long-term tests exist, and why serial retesting in pursuit of a number you like tells you about your patience, not your house. More on sequencing: testing vs. mitigation.
Some of that spread is the house and some of it is the test. A short measurement answers a narrower question than its single number implies, which is why how long the test actually ran is worth checking before you trust it — and the commonest reason a low reading flatters a house is that a window was open while it ran. Both are why reading one result as the house’s permanent value is the interpretation that goes wrong most often. It is also why a result whose conditions were never written down cannot really be compared with a later one: without them there is nothing to hold constant between the two numbers.
What mitigated homes typically reach
EPA’s consumer guidance is that most homes can be reduced below 4 pCi/L, and well-designed systems commonly verify below 2. The proof is always the post-mitigation test — insist on it in any scope of work. See how systems get there.
Common questions
Is 3.9 “safe” and 4.1 “dangerous”?
Neither word fits. Radon risk is a continuous curve — 4 pCi/L is where EPA plants its action flag for policy and practicality, not a boundary in biology. A 3.9 and a 4.1 describe essentially the same house. Read results in bands, decide with the band’s guidance, and spend exactly zero energy on the second decimal.
What does WHO say versus EPA?
The World Health Organization discusses a lower reference level (commonly cited at 2.7 pCi/L) while EPA uses 4 with a 2–4 consideration band. The two agree on the substance: risk rises with concentration, there is no perfectly safe level, and reduction is worthwhile as numbers climb. For U.S. homes and transactions, EPA’s framework is the operating standard — and this site cites it exclusively.
My result came back in becquerels (Bq/m³). How do I convert?
Divide by 37: 148 Bq/m³ equals 4 pCi/L, so a 200 Bq/m³ report is about 5.4 pCi/L — above EPA’s action level. Some labs and international devices default to becquerels; the guidance bands on this page apply identically once converted.
Does an elevated result mean we’ve been harmed already?
It means exposure, not injury — radon risk accumulates over years, which is why discovering a high number after living with it for a while calls for fixing it now rather than panic about the past. If you smoke, the combined risk is substantially higher (EPA is direct about this), making mitigation and quitting resources a powerful pair.
What to do with a radon result
What it shows. That the first question is test validity, not the number — where most disputed results actually resolve.
Authored as SVG in this repository, with real text nodes and a long description for screen readers. Reusable under CC BY 4.0 — see the diagram library.
Your number is in the action band?
Turn it into a priced, scheduled fix — free quotes from independent IEMA-licensed contractors.
Sources
- U.S. EPA — Radon
- U.S. EPA — A Citizen’s Guide to Radon
- U.S. EPA — Health Risk of Radon
- U.S. EPA — Consumer’s Guide to Radon Reduction
- IEMA-OHS — Illinois Emergency Management Agency and Office of Homeland Security — Radon Program