Why national data can tell you how often Canadian homes test high, and never whether yours will
Credit: Homeowner.ca
Almost every question Canadians ask about radon is, underneath, an attempt to avoid testing for it. Is my province one of the bad ones? My house was built in 2019, so it must be fine. My neighbour tested and came back low. The basement is unfinished and we barely go down there.
Each of those is a reasonable thought. None of them is a measurement.
That distinction is the whole subject. Radon research in Canada is unusually good, and it can tell you with real precision how often homes of a given age, in a given region, with a given foundation, come back above the guideline. What it cannot do — what no dataset can do — is tell you which side of that distribution your particular house sits on. Population data describes a spread. You live in one house.
This matters more than it sounds, because the patterns are genuinely counterintuitive. Newer Canadian homes do not test lower than older ones. Homes without basements are not exempt. The energy retrofit you were encouraged to do can move the number in the wrong direction. And several of the beliefs that look most obviously false turn out to be partly true, which is precisely what makes them durable.
So rather than sort these into myths and facts, we have graded each one: false, partly true, or test-dependent. The grade tells you what the evidence supports. The action tells you what to do about your own home, which in nearly every case is the same thing, for reasons that will be clear by the end.
Start with a puzzle, because it explains everything that follows.
Ask how many Canadian homes have high radon and you will find four different answers in circulation, all of them sourced, all of them current enough to be quoted in the press this year. Health Canada's own factsheet now states plainly that 1 in 5 homes in Canada have high radon levels. That is a striking number, and it is the most recent federal framing.
It is also roughly triple what the federal government reported a decade ago. CAREX Canada, the carcinogen surveillance program hosted at the University of British Columbia, records that in Health Canada's randomized cross-country survey conducted from 2009 to 2011, 6.9% of homes tested above the Canadian guideline of 200 becquerels per cubic metre.
A third figure comes from the research literature. When University of Calgary researchers analyzed 11,727 Alberta and Saskatchewan homes tested between 2010 and 2018, 17.8% measured at or above 200 Bq/m³ — but that is two Prairie provinces, not the country.
A fourth comes from the 2024 Cross-Canada Radon Survey, which reports 17.8% of all residential buildings and 20.4% of detached houses at or above the guideline. That survey is citizen science: participants chose to order a test kit. People who suspect a problem are more likely to volunteer, which pushes a self-selected sample upward relative to a randomized one.
None of these numbers is wrong. They measure different populations, by different methods, in different regions, in different decades. Detached houses are not all residential buildings. Alberta and Saskatchewan are not Canada. A self-selected sample is not a random one. Anyone quoting a single national radon statistic without saying which of those they mean is, at best, being imprecise.
Hold onto that, because it is the article's argument in miniature. If four careful Canadian datasets cannot agree on a national figure, the odds that any of them describes your living room are not encouraging. They were never meant to.
Geography is the belief with the strongest grip, and it has real evidence behind it.
Radon originates in soil and rock, so the amount available to enter a building genuinely varies with local geology. CAREX Canada quantifies this as radiation dose: Canadians absorb roughly 1.0 millisievert a year from inhaling radon and its decay products, but in Vancouver the average is 0.2 mSv a year while in Winnipeg it is 2.2 mSv. That is an elevenfold difference between two Canadian cities, and it is not noise.
Regional differences show up in test results too. The 2024 survey found roughly 1 in 3 Atlantic properties and 1 in 3 Pacific Interior and Yukon properties at or above the guideline, against 1 in 6 in Central Canada and 1 in 75 on the Pacific Coast. Saskatchewan, Manitoba, New Brunswick and the Yukon led the earlier federal survey.
So the pattern is real. The problem is resolution.
Radon maps are drawn at the health-region level, and CAREX Canada's own map of the 2009–2011 survey uses buckets as wide as 14% to 40% in a single colour band. A map that groups a 14% neighbourhood with a 40% neighbourhood is not telling you about a street. It is telling you about a territory that might take four hours to drive across. Its authors know this: CAREX states that while indoor radon is more prevalent in some regions, no areas of the country are radon free.
Then there is the scale problem underneath that. Health Canada's measurement guide puts it in one sentence: "Radon can vary widely between homes, even in homes in the same building, or buildings on the same street." Same street. Same building. The variation that matters happens at a scale finer than any map Canada publishes.
Which is why the neighbour question resolves cleanly, and in both directions. The researchers behind the national study note that radon levels differ next door because they depend not only on a property's characteristics but on the behaviour and habits of the people inside it. Your neighbour's low reading is not evidence about your house. Neither is their high one.
If you want the regional picture in detail, our breakdown of radon levels by province sets out where Canada's higher-risk areas are — as context for why you should test, never as a substitute for doing it.
Geography is the belief with the strongest grip, and it has real evidence behind it.
Radon originates in soil and rock, so the amount available to enter a building genuinely varies with local geology. CAREX Canada quantifies this as radiation dose: Canadians absorb roughly 1.0 millisievert a year from inhaling radon and its decay products, but in Vancouver the average is 0.2 mSv a year while in Winnipeg it is 2.2 mSv. That is an elevenfold difference between two Canadian cities, and it is not noise.
Regional differences show up in test results too. The 2024 survey found roughly 1 in 3 Atlantic properties and 1 in 3 Pacific Interior and Yukon properties at or above the guideline, against 1 in 6 in Central Canada and 1 in 75 on the Pacific Coast. Saskatchewan, Manitoba, New Brunswick and the Yukon led the earlier federal survey.
So the pattern is real. The problem is resolution.
Radon maps are drawn at the health-region level, and CAREX Canada's own map of the 2009–2011 survey uses buckets as wide as 14% to 40% in a single colour band. A map that groups a 14% neighbourhood with a 40% neighbourhood is not telling you about a street. It is telling you about a territory that might take four hours to drive across. Its authors know this: CAREX states that while indoor radon is more prevalent in some regions, no areas of the country are radon free.
Then there is the scale problem underneath that. Health Canada's measurement guide puts it in one sentence: "Radon can vary widely between homes, even in homes in the same building, or buildings on the same street." Same street. Same building. The variation that matters happens at a scale finer than any map Canada publishes.
Which is why the neighbour question resolves cleanly, and in both directions. The researchers behind the national study note that radon levels differ next door because they depend not only on a property's characteristics but on the behaviour and habits of the people inside it. Your neighbour's low reading is not evidence about your house. Neither is their high one.
If you want the regional picture in detail, our breakdown of radon levels by province sets out where Canada's higher-risk areas are — as context for why you should test, never as a substitute for doing it.
If geography is too coarse, surely the building is the answer. This is where the evidence gets genuinely interesting, because it inverts the assumption most homeowners start from.
The intuition is that a home built in the last fifteen years, under modern codes, should test lower. Canadian data says the opposite. Researchers comparing 25,489 Canadian properties against 38,596 Swedish ones found that Canadian houses built in the 2010s and 2020s average 131 Bq/m³ against 28 Bq/m³ for Swedish equivalents of the same era — 467% greater — and concluded that the radon provisions in the 2010 national building code had not significantly reduced the underlying risk.
The Swedish comparison is what makes this persuasive. Sweden's numbers fell over the same period, so there is nothing inevitable about newer construction producing more radon. It is a consequence of how we build, not a law of physics.
A federal study tested the same question directly. Health Canada researchers measured 225 homes in Halifax and 171 in Winnipeg over three months of the 2021–2022 heating season, comparing homes built before and after the 2011 provincial adoption of the code. Detached Halifax homes built after 2011 were nearly twice as likely to exceed the sample's geometric mean, at an odds ratio of 1.91. In Winnipeg, there was no significant change at all.
That split is the finding. The same code, the same decade, two Canadian cities, two different outcomes. And the same study tested five other building characteristics — storeys, area in contact with the ground, heating system, water source and ventilation type — and found none of them significantly associated with indoor radon in that sample.
Read that again, because it is the most useful negative result in Canadian radon research. A study that set out to find building predictors mostly did not find them.
Some structural effects are real and measurable. Drawing on long-term tests in 21,818 Canadian homes, a Health Canada scientist reported that homes with a basement average about twice the radon of homes without one, and that bungalows run 46% higher than multi-storey detached houses.
Twice is a meaningful multiplier. It is not a guarantee, and this is exactly where "partly true" earns its keep. Health Canada's position on homes above ground level is carefully constructed: such homes should still be tested, because although they typically have lower levels, they can still exceed the guideline. The department concedes the pattern and refuses the conclusion. That is the correct move, and it is the move this entire article is built on.
Our detailed look at the nine home features that predict higher radon works through what each characteristic contributes and, just as importantly, how much it does not.
If geography is too coarse, surely the building is the answer. This is where the evidence gets genuinely interesting, because it inverts the assumption most homeowners start from.
The intuition is that a home built in the last fifteen years, under modern codes, should test lower. Canadian data says the opposite. Researchers comparing 25,489 Canadian properties against 38,596 Swedish ones found that Canadian houses built in the 2010s and 2020s average 131 Bq/m³ against 28 Bq/m³ for Swedish equivalents of the same era — 467% greater — and concluded that the radon provisions in the 2010 national building code had not significantly reduced the underlying risk.
The Swedish comparison is what makes this persuasive. Sweden's numbers fell over the same period, so there is nothing inevitable about newer construction producing more radon. It is a consequence of how we build, not a law of physics.
A federal study tested the same question directly. Health Canada researchers measured 225 homes in Halifax and 171 in Winnipeg over three months of the 2021–2022 heating season, comparing homes built before and after the 2011 provincial adoption of the code. Detached Halifax homes built after 2011 were nearly twice as likely to exceed the sample's geometric mean, at an odds ratio of 1.91. In Winnipeg, there was no significant change at all.
That split is the finding. The same code, the same decade, two Canadian cities, two different outcomes. And the same study tested five other building characteristics — storeys, area in contact with the ground, heating system, water source and ventilation type — and found none of them significantly associated with indoor radon in that sample.
Read that again, because it is the most useful negative result in Canadian radon research. A study that set out to find building predictors mostly did not find them.
Some structural effects are real and measurable. Drawing on long-term tests in 21,818 Canadian homes, a Health Canada scientist reported that homes with a basement average about twice the radon of homes without one, and that bungalows run 46% higher than multi-storey detached houses.
Twice is a meaningful multiplier. It is not a guarantee, and this is exactly where "partly true" earns its keep. Health Canada's position on homes above ground level is carefully constructed: such homes should still be tested, because although they typically have lower levels, they can still exceed the guideline. The department concedes the pattern and refuses the conclusion. That is the correct move, and it is the move this entire article is built on.
Our detailed look at the nine home features that predict higher radon works through what each characteristic contributes and, just as importantly, how much it does not.
Here is the finding most likely to be new to a Canadian homeowner, and the one with the most awkward policy implications.
Health Canada states that energy retrofits — increasing airtightness, sealing cracks and openings, replacing windows and doors, adding insulation — can reduce air exchange and inadvertently increase radon levels. The department cites a United States Department of Energy assessment finding that reducing energy costs raised radon by 22% on average, alongside single-country studies from Germany, the United Kingdom and Switzerland reporting increases from replacing windows and sealing.
Those are four separate programs in four different housing stocks, and they should not be averaged into a number you apply to your own renovation. What they establish is direction, not magnitude.
The mechanism is easier to picture than the statistics. A Health Canada radiation specialist framed it well in a federal science blog: it has become far more accessible to update the top part of your home — the windows and the doors — than to improve the bottom of it. We have spent two decades getting very good at stopping air from leaving houses, and radon leaves houses the same way heat does.
Which brings up ventilation as a remedy, where the honest answer is a qualified one. A National Research Council field study of occupied Canadian houses measured a median radon reduction of 39% from running a heat recovery ventilator continuously at high fan speed, which sits neatly inside Health Canada's estimate that an HRV might cut levels by 25 to 50 percent in most homes.
But notice the starting point. Those houses began at a median of 175 Bq/m³ — below the guideline. A 39% reduction is genuinely useful there. It does not rescue a house measuring 400, and Health Canada is explicit that ventilation's effectiveness for radon reduction is limited and suited only to situations needing modest reductions.
Ventilation moves the number. It does not set it.
If you complete any renovation that changes your home's structure or ventilation — new windows, air sealing, insulation, a finished basement, a new HVAC system — Health Canada recommends re-testing the next heating season. A result from before the work no longer describes the building.
Our guide to energy retrofit mistakes that raise indoor radon covers how to sequence efficiency work so it does not quietly undo your air quality.
Most disagreements about radon results are actually disagreements about what was measured.
Health Canada's guideline is defined against an average annual radon level. Not a reading, not a weekend, not a worst hour — an annual average. The measurement rules follow from that definition: the department recommends testing over a period of three to twelve months, and requires that a minimum of 91 days of the test fall within the heating season.
Those are two separate conditions, and homeowners routinely satisfy the first while failing the second. A 91-day test run from April through July meets the duration rule and misses the standard entirely. Health Canada's own guidance is that tests with fewer than 91 heating-season days risk underestimating the annual average, and its published answer to someone in exactly that position is to test again.
There is a live scientific argument sitting underneath this, and it deserves to be named rather than smoothed over. The Prairie research discussed earlier found minimal winter-to-summer variation in almost half of the properties it examined, with the rest split between higher-winter and higher-summer homes, and the authors concluded that this challenges the usefulness of seasonal correction factors altogether. The national study now recommends six-month tests for anyone starting in spring or summer.
Neither position changes what you should do. Test long. If you are starting outside the heating season, test longer.
The fastest-growing source of radon anxiety in Canada is a small electronic device on a basement shelf displaying a number above 200 after two days.
That number is not a small version of the annual average. It is a different quantity. Health Canada is unambiguous that the 200 Bq/m³ guideline applies to the average annual level rather than to short-term measurements, and that short-term measurements are not appropriate for informing mitigation decisions. Normal fluctuation means brief excursions above 200 are entirely expected in homes that are fine.
How much do short tests actually diverge? The Calgary researchers ran the cleanest available comparison, placing 704 five-day devices within 10 centimetres of long-term detectors in the same rooms over the same period. The five-day tests predicted their 90-plus-day counterparts 80% of the time, an r² of 0.805, against 0.962 for two long-term tests compared with each other.
That is a real loss of precision, and it is worth stating accurately rather than dramatically — short tests are noisier, not useless. The physical reason is simply that indoor radon is unstable over hours: CAREX Canada notes levels can change by a factor of two or three within 24 hours.
Consumer monitors are not the problem, then. Reading them impatiently is. The Canadian National Radon Proficiency Program advises leaving these devices in place for at least three months to get an accurate representation of your average level, warning that short-term measurements can mislead in either direction, and notes that consumer units cannot be professionally calibrated.
There is one legitimate short-term use, and it is worth knowing: after a mitigation system is installed, a shorter test of a few weeks is an accepted way to confirm the system is holding.
Our breakdown of how accurate consumer radon monitors are in Canada covers which units are C-NRPP tested and how to read their long-term averages properly.
A test result is a snapshot of a specific building, in a specific configuration, occupied by specific people. Change any of those and the snapshot ages.
Health Canada recommends re-testing the next heating season after any renovation affecting structure or ventilation, advises that homes with mitigation systems still be tested when no recent measurement exists, and recommends testing every five years where an active soil depressurization system is installed, to confirm it is still working. A mitigation fan can fail without announcing itself.
The same logic applies to a result inherited with the house. The previous owners had different windows, possibly a different basement, and certainly different habits about opening them.
Most disagreements about radon results are actually disagreements about what was measured.
Health Canada's guideline is defined against an average annual radon level. Not a reading, not a weekend, not a worst hour — an annual average. The measurement rules follow from that definition: the department recommends testing over a period of three to twelve months, and requires that a minimum of 91 days of the test fall within the heating season.
Those are two separate conditions, and homeowners routinely satisfy the first while failing the second. A 91-day test run from April through July meets the duration rule and misses the standard entirely. Health Canada's own guidance is that tests with fewer than 91 heating-season days risk underestimating the annual average, and its published answer to someone in exactly that position is to test again.
There is a live scientific argument sitting underneath this, and it deserves to be named rather than smoothed over. The Prairie research discussed earlier found minimal winter-to-summer variation in almost half of the properties it examined, with the rest split between higher-winter and higher-summer homes, and the authors concluded that this challenges the usefulness of seasonal correction factors altogether. The national study now recommends six-month tests for anyone starting in spring or summer.
Neither position changes what you should do. Test long. If you are starting outside the heating season, test longer.
The fastest-growing source of radon anxiety in Canada is a small electronic device on a basement shelf displaying a number above 200 after two days.
That number is not a small version of the annual average. It is a different quantity. Health Canada is unambiguous that the 200 Bq/m³ guideline applies to the average annual level rather than to short-term measurements, and that short-term measurements are not appropriate for informing mitigation decisions. Normal fluctuation means brief excursions above 200 are entirely expected in homes that are fine.
How much do short tests actually diverge? The Calgary researchers ran the cleanest available comparison, placing 704 five-day devices within 10 centimetres of long-term detectors in the same rooms over the same period. The five-day tests predicted their 90-plus-day counterparts 80% of the time, an r² of 0.805, against 0.962 for two long-term tests compared with each other.
That is a real loss of precision, and it is worth stating accurately rather than dramatically — short tests are noisier, not useless. The physical reason is simply that indoor radon is unstable over hours: CAREX Canada notes levels can change by a factor of two or three within 24 hours.
Consumer monitors are not the problem, then. Reading them impatiently is. The Canadian National Radon Proficiency Program advises leaving these devices in place for at least three months to get an accurate representation of your average level, warning that short-term measurements can mislead in either direction, and notes that consumer units cannot be professionally calibrated.
There is one legitimate short-term use, and it is worth knowing: after a mitigation system is installed, a shorter test of a few weeks is an accepted way to confirm the system is holding.
Our breakdown of how accurate consumer radon monitors are in Canada covers which units are C-NRPP tested and how to read their long-term averages properly.
A test result is a snapshot of a specific building, in a specific configuration, occupied by specific people. Change any of those and the snapshot ages.
Health Canada recommends re-testing the next heating season after any renovation affecting structure or ventilation, advises that homes with mitigation systems still be tested when no recent measurement exists, and recommends testing every five years where an active soil depressurization system is installed, to confirm it is still working. A mitigation fan can fail without announcing itself.
The same logic applies to a result inherited with the house. The previous owners had different windows, possibly a different basement, and certainly different habits about opening them.
The single most consequential misconception about radon is that 200 is a boundary between safe and unsafe. It is not. It is a policy judgement, and understanding what kind of object it is changes how you read your own result.
Take Action on Radon, the national awareness initiative delivered with Canadian radon scientists, publishes lifetime lung cancer risk by exposure band. The numbers describe a slope, not a cliff:
Nothing discontinuous happens at 200. Risk rises steadily across the whole range, and the initiative states directly that while risk below the guideline is small, there is no level considered risk free, and that each homeowner decides what exposure they are willing to accept. Researchers make the same point more bluntly: your body cannot distinguish 195 from 205.
That the threshold is a choice becomes obvious once you notice how many different numbers apply to Canadians. Canada's guideline was set at 200 in 2007, down from 800. The World Health Organization recommends a national reference level of 100 and advises countries not to exceed 300. The Canada Labour Code still uses 800 for federal workplaces. And Health Canada told CBC News in March 2026 that it has no plans to change the guideline for now, preferring to concentrate on getting more Canadians to test at all.
What the guideline does mark is where action is recommended, and the current instruction for homes is tighter than the one most Canadian articles still repeat. Above 200 Bq/m³, Health Canada's residential guide calls for corrective action within one year — not the two-year window that survives in its separate guide for public buildings.
On the risk itself, the federal framing is worth reproducing carefully, because it is neither alarming nor reassuring. Radon is the number one cause of lung cancer in non-smokers, responsible for an estimated 16% of lung cancers and more than 3,000 deaths in Canada each year. The risk is also, as Health Canada puts it, long-term rather than immediate, and it compounds sharply with smoking.
That last point is the one to sit with. A high reading is a serious finding that you have time to address properly. It is not an emergency requiring a decision this week — which is precisely why running a proper 91-day test instead of acting on a two-day reading costs you nothing.
For a fuller treatment of what to do at each level, see our guide to understanding radon test results across low, borderline and high readings.
Here is the whole argument in one place. The grade describes what the evidence supports about homes in general. The action describes what to do about yours.
Read down the action column and one thing stands out. Whatever the grade, the answer is the same.
That is not a rhetorical trick. It is the practical consequence of the fact that every row in the evidence column describes a population and every row in the action column describes a building. Health Canada reduces it to a single sentence: no matter the age, type of construction or where your home is located, the only way to be sure of the radon level in your home is to test.
For the full picture across testing, interpretation and reduction, our complete homeowner's guide to radon in Canadian homes pulls the whole subject together in one place.
The practical path is short, and it does not require you to resolve any of the debates above.
Get a long-term alpha-track kit or a C-NRPP-tested electronic monitor, and place it on the lowest level of the home where people normally spend time — a bedroom, office, living room or playroom, not a kitchen, bathroom or laundry room. Start it so that at least 91 days land in the heating season. Then live normally, because a test taken during an unrepresentative few months describes a house nobody lives in.
When the result comes back, read it against the slope rather than the line. Below 200, note the number and plan to re-test within five years. Above 200, Health Canada's current residential guidance is to act within a year, and it recommends using a professional certified through the Canadian National Radon Proficiency Program. Near 200 either way, the decision is legitimately yours, and the factors that should move it are who lives in the house, how many hours they spend in it, and whether anyone smokes or has smoked.
If you are buying or selling, a radon result is a negotiable item like any other inspection finding, and a previously installed mitigation system is generally read by the market as a feature rather than a defect.
Our overview of radon testing for home sales in Canada sets out what buyers and sellers should reasonably ask for, and what a rushed test during a closing window can and cannot establish.
About the Author
Ryan is the founder of Homeowner.ca and a proud Canadian homeowner based in Guelph, Ontario. Over his 25-year career in digital publishing, he has focused on transforming complex information into clear, practical guidance that helps people make confident, well-informed decisions.
Yes. Regional averages are real but are published at health-region scale, which is far coarser than the scale at which radon actually varies. Health Canada notes that levels differ widely between homes on the same street, and no region of Canada is radon-free. The map tells you about a territory; only a test tells you about a house.
No, and this holds in both directions. Radon levels next door depend on that building's construction details and on how its occupants live — how often windows open, how the basement is used, how the ventilation runs. A low neighbouring result is not reassurance, and a high one is not a diagnosis.
Not reliably. Health Canada has found that the rough-in stub on its own was generally insufficient, and that passive systems cannot be relied on to bring high concentrations below the guideline. Canadian homes built in the last fifteen years also tend to test higher, not lower, than older stock. Test regardless.
No. Homes with basements do average roughly twice the radon of homes without, so the belief is not baseless. But Health Canada is clear that homes above ground level should still be tested, because they can still exceed the guideline. Test the lowest level where you spend time.
Not yet. The guideline applies to an annual average, and Health Canada states that short-term measurements are not appropriate for informing mitigation decisions. Indoor radon can change by a factor of two or three within a single day, so brief excursions above 200 occur in homes whose annual average is fine. Leave the device running to at least 91 days and read the long-term average.
In the cleanest Canadian comparison, five-day tests placed alongside long-term detectors in the same rooms predicted their 90-day counterparts about 80% of the time, against 96% agreement between two long-term tests. That is meaningfully noisier, not worthless — which is why short tests are accepted for confirming that a newly installed mitigation system is working, but not for deciding whether to install one.
Treat it as history rather than data. Their result reflects the building as it was configured and occupied then. Health Canada recommends testing when no recent measurement is available, re-testing after any renovation affecting structure or ventilation, and testing every five years where a mitigation system is installed.
Both, ideally, but the essential one is after. Health Canada notes that energy retrofits can reduce air exchange and inadvertently increase radon, and recommends re-testing the next heating season following work that changes structure or ventilation. A pre-renovation result does not describe the post-renovation building.
It is below the Canadian guideline, which is not the same as safe. Risk rises continuously with exposure and no level is considered risk free. At a borderline reading the decision is genuinely yours, and it should weigh who lives in the home, how many hours a day they spend there, and whether anyone smokes or used to.
Because a guideline is a policy judgement balancing health risk against practicality, not a biological threshold. Canada set 200 in 2007, down from 800; the WHO recommends 100 and advises countries not to exceed 300; the Canada Labour Code still uses 800 for federal workplaces. Health Canada said in early 2026 that it has no plans to change the number, preferring to focus on increasing how many Canadians test.
Health Canada's standard requires at least 91 days of the test to fall in the heating season, and warns that tests with fewer heating-season days risk underestimating the annual average. Canadian research has found minimal winter-to-summer variation in nearly half the homes studied, which complicates the seasonal assumption but does not change the standard. If you start outside the heating season, run the test for six months or longer.
