Is my wood burner damaging my health? I tracked pollution in my home to find out

Recently, I’ve been increasingly concerned about the impact of wood burning on my health.
My parents live in rural Cornwall, and when I visit them in the winter months, we typically light their wood burner every evening.
However, a growing body of evidence suggests that wood burning is a key source of a type of air pollution called particulate matter (PM), which has been linked to conditions ranging from dementia to stomach cancer.
To find out whether our stove use could be negatively impacting my family's health, I used a monitor to track air pollution in my parents' living room for a week. My findings broadly aligned with existing literature and inspired us to change the way we use the stove.
For a detailed look at the health impacts, see our guide to the pollution and health impacts of wood-burning stoves
What is particulate matter?
It's long been established that wood burning creates particulate matter (PM) – tiny solid particles that are suspended in the air. These include dirt, dust, and soot.
Research into the health impacts of wood-burning stoves has focused on PM2.5 and PM10. ‘2.5’ and ‘10’ refer to the diameter, in micrometres, of the particles making up the particulate matter. For context, the average human hair is about 70 micrometres in diameter.
Due to the size of these particles, they are small enough to make their way into your lungs when they are inhaled, and then, in some cases, cross into your bloodstream. This means they can be carried around the body and reach other organs.
A 2019 review of research linked PM exposure to conditions ranging from heart disease to bladder cancer. In 2025, the University of Cambridge published a report linking long-term exposure to air pollution to dementia. It suggests that exposure to air pollution can be linked to inflammation and oxidative stress in the body and brain, both of which play a well-established role in the onset and progression of dementia.
Tracking my PM exposure while using a wood-burning stove
At the end of February 2026, I used the wood-burning stove at my parents' home in Cornwall every evening for five nights, for several hours.
To track air pollution, I ran an air quality monitor continuously in the living room near the stove for the entire study period. It measured the level of PM2.5 in the room every two minutes. I noted down whenever the stove was lit and whenever more wood was added.
Pollutants are typically measured in micrograms per cubic metre (µg/m3). This shows the concentration of a particular pollutant in the air. For example, a concentration of 1µg/m3 means that every one cubic metre of air contains one microgram (one millionth of a gram) of that particle.
Once the study period was finished, I was able to map spikes in air pollution in the home. Read on to see my key findings.
Find out about wood-burning stove fines and regulations
1) Air pollution was higher when the stove was in use
To see whether PM2.5 concentrations were higher when the stove was lit, I averaged readings from the air pollution monitor during the times when the stove was in use (‘burning period’) and when it wasn’t (‘non-burning period’).
Average PM2.5 concentrations during burning and non-burning periods
| Day 1 | Day 2 | Day 3 | Day 4 | Day 5 | |
|---|---|---|---|---|---|
| PM2.5 – Burning period (µg/m3) | 4.91 | 11.80 | 4.98 | 11.43 | 11.94 |
| PM2.5 – Non-burning period (µg/m3) | 0.01 | 0.67 | 0.20 | 0.63 | 1.74 |
Table notes: Based on PM2.5 readings collected between 27 February 2026 and 03 March 2026. Burning periods ran from the time the stove was lit to three hours past the last time fuel was added. Where burning periods extended past midnight, I included those readings in the previous days' burning period and not the day they were taken – see 'Study details' for timings.
During non-burning periods, recorded levels of air pollution were consistently low – average PM concentrations never exceeded 2.5 µg/m3. This is typical for a rural area.
In contrast, average PM concentrations were greater than 10µg/m3 during three of the five burning periods.
Looking across all days, the average PM concentration during non-burning periods was 0.64 µg/m3, and during burning periods was 9.34 µg/m3, a statistically significant difference.
2) I observed spikes in air pollution after the stove was lit
During my analysis, I noticed that air pollution rose sharply after the stove was lit before gradually tapering back down.
The graph below shows an example of PM2.5 concentrations over the course of an evening.
Average hourly PM2.5 concentrations between 12pm and 11.59pm on 1 March 2026
Table notes: Based on average hourly PM2.5 readings collected on 1 March 2026. Each data point corresponds to the average for the following hour. For example, the data point at 7pm is the average of readings taken between 7 and 8pm.
On this day, before lighting the stove, PM2.5 concentrations ranged from roughly 0µg/m3 to 1µg/m3.
I lit the stove at 7.10pm. In the hour following, between 8pm and 9pm, the average PM2.5 concentration was 16.24µg/m3, and between 9pm and 10pm it was 10.07µg/m3.
I added one log to the stove at 11pm, when it was smouldering at a low level. PM levels remained raised, but did not spike again.
3) Spikes in pollution were significantly greater than average levels of pollution
To determine if the spikes in pollution after the stove was lit were significantly greater than average levels of pollution, I replicated the statistical analysis used in a paper by Chakraborty et al (2020).
Their study, which tracked pollution across four homes with wood burners, found statistically significant spikes in pollution during stove use.
To replicate their analysis, I calculated the average concentrations of PM2.5 for each hour across all the study days. Then I identified the hour in each day with the greatest concentration (called the ‘peak hourly average’) and averaged across all the hourly averages in each day to find the ‘daily hourly average’.
Peak and daily hourly averages of PM2.5
| 27/02 | 28/02 | 01/03 | 02/03 | 03/03 | |
|---|---|---|---|---|---|
| PM2.5 – Peak hourly average (µg/m3) | 9.13 | 24.27 | 16.24 | 19.49 | 24.46 |
| PM2.5 – Daily hourly average (µg/m3) | 0.62 | 3.31 | 1.51 | 3.59 | 4.20 |
Table notes: Based on PM2.5 readings collected between 27 February 2026 and 04 March 2026. Burning periods ran from the time the stove was lit to three hours past the last time fuel was added. Where burning periods extended past midnight, I included those readings in the previous days' burning period and not the day they were taken – see 'Study details' for timings.
While daily hourly averages were never greater than 5µg/m3, peak hourly averages reached more than 15µg/m3 on four of the five days.
All of these peaks occurred while the stove was lit.
The highest concentration of PM2.5 was on 3 March, when levels reached 24.46µg/m3.
I checked if these differences were statistically significant using confidence intervals. I found that the mean peak hourly average of PM2.5 was significantly higher than the mean daily hourly average.
When we report 95% confidence intervals (CIs) it means the "true" or "real" average is likely contained between those values.
CIs are calculated by working out the sample mean, then adding (for the upper CI) or subtracting (for the lower CI) a margin of error, which depends on the variability in the data and the sample size.
The mean peak hourly average was 18.72µg/m3, and the 95% CIs were 10.80 and 26.63. The mean daily hourly average was 2.65µg/m3, and the 95% CIs were 0.77 and 4.53.
Because the CIs for the peak and daily hourly averages don't overlap, this provides strong evidence that the "true" averages differ.
4) My results are replicated in scientific literature
Multiple published studies have reported similar findings to those in my investigation.
Note that the results below include analysis of levels of PM2.5 and PM10. PM10 includes PM molecules up to 10 micrometres in diameter (so measurements include PM2.5 pollution).
Authors, years and key findings from three major studies of wood burning stoves
| Researchers | Year | Findings |
|---|---|---|
| Chakraborty et al | 2020 | Across 4 homes, observed a 196.23% increase in PM2.5 concentrations when stoves were in use. |
| Vicente et al | 2020 | Examined one stove in an unoccupied rural home and found 100% increase in PM10 during stove operation. |
| Kuye and Kumar | 2025 | Across four homes with Ecodesign stoves, found a 200% increase in PM10 during stove operation. |
Table notes: Collected 18/03/2025. These summaries reflect only the key findings from each study, and not the full results.
There is some variation in the results reported - in particular, PM2.5 concentrations during non-burning periods were a lot lower in my study, which means the increases I saw were more dramatic.
However, like my investigation, all the above studies reported statistically significant increases in PM exposure during burning.
Is my wood burner impacting my health?
Given my air monitor assessment, it appears that my family's wood-burning stove does cause pollution spikes during use.
But the key question is: does this pose a danger to our health?
According to the World Health Organization, the recommended limit for average PM2. 5 exposure over a 24-hour period is 15µg/m3. The daily averages in this study fall comfortably below these guidelines.
However, both the UK government and the European Public Health Alliance report that there is no evidence of a safe level of air pollution (i.e., an amount of exposure that will not cause adverse health effects).
To find out more, I spoke with Professor James Allan, professor of Air Pollution Measurement at the University of Manchester, and Dr James Heydon, associate professor at the School of Sociology and Social Policy at the University of Nottingham.
'Even studies of relatively low levels of air pollution have found adverse health effects,' said Professor Allan. 'In addition, it seems that particulate matter from wood burning is a particularly harmful form of pollutant.'
Dr Heydon concurred: 'Research suggests that PM2.5 is especially damaging to your health because it can enter your bloodstream, which means it can cause inflammatory responses throughout the body.'
Growing evidence suggests that both long-term exposure to low levels of PM and brief exposure to high levels of PM can have detrimental health effects.
A 2021 study found exposure to an annual average of less than 10µg/m3 of PM2.5 was associated with an increased risk of mortality. Three months of exposure to similar levels of pollution induced neuroinflammation and cell death in mice.
In addition, a 2012 study found that people exposed to bigger hourly peaks of PM2.5 had an increased risk of mortality – even after controlling for average daily exposure.
Dr Heydon said: 'Even small amounts of air pollution can detrimentally impact your health. Decreasing exposure will yield health benefits across the board.'
Although the exact amount of pollution from my family's stove (and, by extension, the specific health impact of my stove) isn't completely clear, we've decided we won't continue to use it every single night, as we don't feel it's worth the potential health risks of even low levels of daily PM exposure.
We won't stop using it altogether, but will consider it to be a treat for special occasions rather than something to do every day.
See our advice on gas and electric fires
What does this mean for your wood burner?

It's difficult to generalise results across studies of wood burners. Scientific studies find varying levels of pollution during stove usage across homes. I found that PM concentrations in my home varied from day to day.
Dr Allan explained: 'The amount of indoor air pollution produced by wood burning stoves can vary a lot. It depends on factors such as the type of stove, the way it's used, and the level of ventilation in the home.'
Dr Heydon concurred: 'Even activities like cooking and cleaning can influence indoor air pollution, which makes it harder to interpret results. But studies have consistently identified intense bursts of pollution during stove use.'
The key takeaway from the scientific literature is that stoves likely emit some PM pollution during normal use, and that this can contribute to a variety of health conditions.
When considering your stove usage, it's important to balance your enjoyment of using your stove and your heating needs with the potential health impacts of wood burning.
How to reduce the air pollution impacts of a wood-burning stove
There are several ways you can reduce the amount of pollution your stove produces when you do want to enjoy it:
- Don't skip regular servicing and chimney sweeps You should get your chimney swept and your stove serviced at least once a year.
- Burn seasoned or kiln-dried wood Analysis by researchers at the University of Surrey has found that manufactured fuels (for example, smokeless coal and wood briquettes) release significantly more pollution than these types of wood fuel.
- Ventilate the room during use Leaving a door open or intermittently opening a door while your stove is burning will help to disperse pollution.
- Use the top-down technique for lighting your stove 'Stacking your kindling and firelighter on top of your logs appears to help reduce pollution during lighting,' said Professor Allan.
- Use proper refuelling technique When you need to add more fuel to your stove, open the door slightly, pause for a few seconds, then open it all the way. This helps to prevent pollution ‘flooding’ into the room. If you can, you should also wait until the wood in your stove is smouldering, rather than actively burning.
- Consider updating very old stoves If you do have a very old stove, updating to a newer model might help to reduce pollution.
However, it's important not to overstate the impact of these interventions.
'Proper stove use will reduce pollution overall, but it doesn't guarantee that you will not be exposed to high levels of PM during burning,' said Dr Heydon. 'There's no way to control for all the factors that influence air pollution from your stove.'
Find out more about the best heating for your home
More details of my wood-burning stove pollution study
Limitations of the study
It's important to note some of the key limitations of this study when interpreting the results.
- Factors such as humidity and outdoor air quality can also influence pollution levels and the accuracy of monitoring. These factors weren't controlled for during data collection and analysis, which may have led to inaccurate results.
- These results are based on just five days of monitoring in one home, so aren't generalisable beyond that time period or location.
How I used the stove
The stove was used every day during the recording period. All doors and windows were kept shut during this time.
The stove was used in the same way on all recording days: it was initially loaded with approximately eight logs, and half a firelighter was used as kindling.
The stove was refuelled in accordance with HETAS' recommendations, and only one log was added during refuelling.
Stove and room details
| Stove location | Living room |
| Room size (LxWxH) | 6.8 x 4.5 x 2.1 metres |
| Fuel type | Kiln-dried logs |
| Sampling period | 26 February 2026 to 04 March 2026 |
| Stove type | Defra-exempt, assumed not to be Ecodesign-compliant (as it's an old stove that's now discontinued) |
| Stove purchase date | February 2010 |
| Date of last chimney sweep and stove service | December 2024 |
Burning period dates and timings
| Date | Non-burning period times | Burning period times | Number refuels | Total burn time (minutes) |
|---|---|---|---|---|
| 27 February | 0:00-21:05 | 21:06-00:05 | 0 | 180 |
| 28 February | 00:06-18:40 | 18:41-00:30 | 1 | 350 |
| 1 March | 00:31-19:10 | 19:11-02:00 | 1 | 410 |
| 2 March | 02:01-18:50 | 18:51-01:15 | 1 | 385 |
| 3 March | 01:16-20:00 | 20:01-01:50 | 1 | 350 |
I defined a burning period as the time at which the stove was lit until three hours after the last log was added. I chose this endpoint because I refuelled on average three hours after lighting the stove, suggesting that this is when it begins to cool noticeably.
When the burning period ran past midnight, I included those measurements in the previous day's burning period. For example, I included the measurements taken between 00:00 and 00:05 on 28/02 in the burning period for 27 February.
This was to avoid excluding the end of the burning period on 3 March, when pollution was lower, which would have increased the mean burning period on that day.
How I collected and calculated pollution measurements
The air quality monitor was run continuously from 7pm 26 February to 11am 4 March. It took measurements every two minutes. I disregarded data taken before 00:00 on 27 February, and after 01:51 on 4 March.
There were no missing measurements across the whole study period.
To calculate the burning and non-burning period means
To compare pollution levels when the stove was burning and not burning, I averaged readings across these two periods for each day of the study. I chose to calculate the mean PM2.5 level in order to capture the impact of hourly spikes.
I also calculated the mean PM2.5 concentration across all non-burning and burning periods, and ran a Welch's t-test to assess a significant difference between these values. Results were significant (non-burning period mean: 0.64µg/m3; burning period mean: 9.34µg/m3)(Welch’s t(850.44) = −30.69, p < 0.0001)
To calculate the peak hourly average and the daily hourly average concentration
I calculated mean hourly concentrations for all hours of the study.
I calculated the peak hourly average by identifying the hour with the largest reading.
To calculate the daily hourly average, I averaged across all hourly concentrations for the day (including the peak hour).
I rounded burning periods to the nearest whole hour (for example, if the stove stopped burning at 00:30, the last burning hour was considered to be 00:00-01:00). Per the above analysis, when the burning period ran past midnight, those measurements were included in the previous day's burning period.
I also calculated the mean peak hourly average and the daily hourly average across all days, along with 95% confidence intervals for these values. As confidence intervals didn't overlap, this was taken to indicate a statistically significant difference.


