You put a pan on high, the oil shimmers, and within two minutes there is a visible haze at eye level. The smoke alarm in the hallway is thinking about it. The fan above the stove stays off, because it is loud and dinner is almost done.
That is an everyday indoor air exposure, and almost nobody treats it as one.
From 2015 to 2021, about a third of U.S. households, 33.3 percent, used piped gas for cooking, according to American Housing Survey data cited by the Government Accountability Office. Gas stoves have been argued about loudly for several years, and the argument has produced more heat than information. That is a shame, because the practical findings underneath it are useful, and they apply to electric and induction kitchens too.
This article reports what federal investigators and peer-reviewed researchers have found, including where they disagree. It is not here to tell you what appliance to own. EPA’s ranking for any indoor air problem puts the source first, fresh air second and filtration third. The stove is a source, so the fixes that matter happen at the stove.
What the federal review of gas stoves found
In March 2025, the Government Accountability Office published Gas Stoves: Risks and Safety Standards Related to Products and Ventilation, report number GAO-25-107514. Congress asked for it. GAO is non-partisan and has no regulatory authority of its own, which makes it an unusually clean place to start on a contested subject.
How GAO built the report defines what it can and cannot settle. Between April 2024 and March 2025, GAO reviewed the published literature and interviewed the Consumer Product Safety Commission, EPA, the Department of Energy, the Department of Housing and Urban Development, standards organizations including the CSA Group and ASHRAE, testing bodies, industry associations, two appliance manufacturers and two academic researchers.
GAO’s finding has two halves, and both are the finding. Gas stoves carry health and safety risks, partly from emissions such as nitrogen dioxide. And, in the Government Accountability Office’s own words, “there is ongoing debate about the extent of their impact on human health.”
GAO also states the risk is reducible rather than fixed. The report puts it this way: “Potentially harmful emissions from gas stoves can be mitigated through ventilation that directs emissions to the outdoors.”
Reporting one half and not the other is editing, not summarizing. This article keeps both in view.
Stove safety runs through voluntary industry consensus standards, third-party testing and local building codes, with CPSC taking part in standards work and handling recalls. GAO flags one wrinkle that explains a lot about real kitchens: stove standards and ventilation standards come from different organizations, so emissions and the exhaust meant to capture them are not designed together.
The pollutants at issue
GAO narrowed its focus to three substances because EPA identifies them as the major pollutants from gas stoves: nitrogen dioxide, particulate matter and carbon monoxide. Gas cooking also releases carbon dioxide, methane and benzene, among other gases.
Nitrogen dioxide is a respiratory irritant. GAO notes that some studies have measured indoor levels above EPA’s one-hour outdoor limit and above World Health Organization recommendations, particularly in small homes or enclosed kitchens. Room size matters a great deal here, and it comes up again below.
Particulate matter from a gas stove includes fine particles smaller than 2.5 micrometers, including during pan frying, plus ultrafine particles of 0.1 micrometers or less from combustion. GAO notes these may affect heart and lung health. Carbon monoxide gets its own section, because its risk profile is different in kind.
Benzene has the sharpest measurements of the group. Writing in Environmental Science and Technology in 2023, Kashtan and colleagues sampled 87 homes in California and Colorado. Gas and propane burners on high, and ovens set to 350 degrees Fahrenheit, emitted benzene averaging 2.8 to 6.5 micrograms per minute, roughly 10 to 25 times the emissions from electric coil and radiant alternatives. Induction stoves emitted none that was detectable, and neither did the food.
Read that carefully. It measures what comes out of an appliance. It does not measure how many people get sick, and the researchers did not claim it did.
Where the researchers agree, and where they do not
The most ambitious recent U.S. estimate ran in Science Advances in May 2024. A team led by Yannai Kashtan at Stanford, with PSE Healthy Energy and Harvard’s school of public health, combined measurements from more than 100 homes with a room-level indoor air model and national housing data.
Their estimate: gas and propane stoves raise long-term nitrogen dioxide exposure by about 4.0 parts per billion on average nationally, roughly 75 percent of the World Health Organization’s exposure guideline. They attributed about 50,000 cases of current pediatric asthma to long-term stove-related nitrogen dioxide alone.
The finding most readers miss is about square footage. People in homes under 800 square feet took on about four times the long-term exposure of people in homes over 3,000 square feet. The same appliance in a smaller kitchen is a different exposure. That is a housing finding as much as an appliance finding.
Now the disagreement, stated straight.
In 2013, Weiwei Lin, Bert Brunekreef and Ulrike Gehring published a meta-analysis in the International Journal of Epidemiology pooling 41 population studies. They reported a summary odds ratio of 1.32 for childhood asthma and gas cooking exposure, with a 95 percent confidence interval of 1.18 to 1.48, and concluded that gas cooking increases asthma risk in children.
In 2023, Wenchao Li and colleagues published a systematic review in Global Epidemiology covering 66 studies, including the ones in that meta-analysis. They concluded that the epidemiological literature does not provide sufficient evidence of a causal relationship between gas cooking and asthma or wheeze. Their reasons were specific: most studies are cross-sectional, so you cannot tell which came first; exposure and symptoms are usually self-reported; and many studies adjust poorly for secondhand smoke, family history of asthma or allergy, and socioeconomic status.
Here is the disclosure that belongs in the same breath. All six authors were employed by Gradient, a private environmental consulting firm, and the American Gas Association funded the work. The authors state that the association was not involved in drafting and that they had sole responsibility for the content.
Funding does not settle whether an argument is right. It tells you to read the paper alongside independent work rather than instead of it, and to check whether the criticisms hold up on their own. Several of these are the standard objections epidemiologists raise about cross-sectional data, and they are checkable.
GAO landed in the middle, which is where the evidence is. The report states there is no consensus on the health effects of nitrogen dioxide emissions directly attributable to gas stove cooking, and it describes both the 2013 meta-analysis and the later industry-funded review.
What an odds ratio is not
One piece of plain arithmetic makes this whole dispute easier to read.
An odds ratio of 1.32 compares two groups. It does not tell any individual their chance of anything. If a condition is uncommon to begin with, a 32 percent relative increase is a small absolute change. If it is common, the same ratio means many more actual cases. The number alone does not say which situation you are in.
Confounding is the other idea worth owning. If households with gas stoves also differ in housing age, ventilation, income or smoking, a study that does not account for those differences can credit the stove with an effect belonging to something else. That is the ordinary hard problem of studying people in their own homes, not a conspiracy theory.
Carbon monoxide is a different kind of risk
Carbon monoxide has no color and no smell, and it is the one gas here with an emergency plan attached.
Start with the reassuring number. EPA reports that levels near properly adjusted gas stoves are often 5 to 15 parts per million, and that poorly adjusted stoves can produce 30 parts per million or more. GAO’s version matches: carbon monoxide is usually not hazardous in normal use, but dangerous levels can occur when a stove malfunctions or is used incorrectly where ventilation is inadequate. The hazard is a broken or misused appliance, not a working one.
Broken appliances exist. GAO counted 35 gas stove recalls announced by the Consumer Product Safety Commission between June 1973 and February 2024, and about 60,000 stoves recalled in 2022 and 2023 specifically over dangerous carbon monoxide levels.
The national picture, per CDC: each year more than 400 Americans die from unintentional carbon monoxide poisoning not linked to fires, more than 100,000 go to an emergency department, and more than 14,000 are hospitalized.
Symptoms are the problem. CDC lists headache, dizziness, weakness, upset stomach, vomiting, chest pain and confusion, and notes they are often flu-like. EPA adds the detail that gives it away: flu-like symptoms that clear up after leaving the home. Sick at home, better outside, sick again on returning. Say exactly that to a clinician. People sleeping or intoxicated can be poisoned before any symptom registers.
CDC’s alarm guidance is specific. Put battery-operated or battery-backup carbon monoxide detectors near every sleeping area. Check the batteries when you change clocks in spring and fall. Replace the detector following the manufacturer’s instructions or every five years. CDC also suggests considering a model with a digital readout, which shows the highest level recorded in the home.
If a carbon monoxide alarm sounds, get everyone out into fresh air and do not go back in to investigate. Emergency: 911. If anyone is unconscious, confused or will not wake, that is 911 immediately.
One more item here: a gas oven or burner is not a space heater. Heating a room that way runs a combustion appliance for hours without the limits cooking normally puts on it.
Every kitchen makes particles, whatever the stove runs on
This is the part of the subject the public argument has almost entirely buried.
GAO says it directly: fires, burns and particulate emissions from certain cooking methods affect both gas and electric stoves. Searing, frying and high-heat roasting put fine particles into the air because of what is happening to the food and the oil, not because of what heats the pan.
The measurements back that up. In Environmental Science and Technology in 2008, Lance Wallace and colleagues at the National Institute of Standards and Technology ran more than 150 tests in an instrumented test house, using a gas stove, an electric stove and an electric toaster oven. All three generated ultrafine particles, peaking at a particle size of roughly 5 nanometers. Total counts ran as much as 10 times higher than earlier studies had reported, because older instruments could not see particles that small.
The same study found that less than 5 percent of outdoor particles under 10 nanometers made it into the house. When those particles are in your kitchen, they came from your kitchen.
Cooking shows up in exposure terms too. Sun and Wallace, in the Journal of the Air and Waste Management Association in 2021, tracked 132 households in Halifax and Edmonton. Cooking accounted for about 22 percent of residents’ total daily fine particle exposure. Those were Canadian homes, so treat the share as indicative rather than an American figure.
EPA’s own guidance reflects all this. Its page on indoor particulate matter names cooking as a source and tells people to install and use exhaust fans vented to the outside when they cook. Nothing in that sentence mentions fuel.
What that does to a body is a thinner line of work. Gabdrashova and colleagues, in Science of the Total Environment in 2020, had 50 healthy volunteers fry ground beef on an electric stove and found ultrafine particle exposure associated with a statistically significant rise in systolic blood pressure. One small controlled study is preliminary evidence, not an established effect.
A self-cleaning oven cycle is the extreme version of high-heat cooking: hours of burning food residue off the oven walls. No federal measurement of what that cycle releases turned up in the sources available for this article, so treat it as a reason for the same precaution rather than a quantified risk. Run the hood, open a window, and give the room time afterward.
How range hoods perform
Kitchen exhaust is the intervention with the most evidence behind it, and the one most households skip.
GAO is candid that hoods vary. Performance depends on fan airflow, on how fully the hood covers the burners, on kitchen and home size, and on how the hood interacts with whole-house ventilation. A hood covering all the burners typically captures more. Because effectiveness varies, some emissions stay in the room or move into the rest of the home.
The single most useful detail in the GAO report concerns burner position: according to some researchers, hoods capture more emissions from back burners than from front burners. That is free. It costs nothing to slide the stockpot to the back.
The second most useful detail is about your own kitchen, and you can check it today. The International Residential Code, developed by the International Code Council and adopted in most states with local modifications, permits recirculating range hoods alongside whole-house ventilation. A recirculating hood pulls air through a grease filter and blows it back into the room. It captures grease. It does not remove gases, and it removes nothing from the house. Over-the-range microwaves are often installed in recirculating mode even when they could be ducted.
How to tell: look for a duct above the hood or behind the microwave, and watch where the air comes out. If warm air blows back into the kitchen from a vent at the top front of the unit, it is recirculating.
Behavior is the other half. In those 132 Canadian households, only 27 percent of cooking activities involved any added ventilation. Range hoods were used in 10 percent of cooking events, windows opened in 15 percent, both in 2 percent. Using a hood or opening a kitchen window doubled the rate at which fine particles cleared from the air.
The most encouraging result in this field comes from a small trial. Stephanie Holm and colleagues at the University of California, San Francisco and Lawrence Berkeley National Laboratory published it in the Journal of Exposure Science and Environmental Epidemiology in 2024. They worked with 14 Bay Area homes that had a gas stove and a school-aged child. Midway through monitoring, families watched a four-minute video asking them to use the back burners and run the hood every time they cooked. Researchers also replaced hoods that did not meet airflow and sound standards.
Back burner use rose substantially. Hood use rose slightly. Cooking frequency and duration did not change at all. Nitrogen dioxide fell: integrated concentration across cooking events dropped from 1,230 parts per billion-minutes to 756, and nitrogen dioxide on samplers across the whole period fell from 10.4 to 9.4 parts per billion. Fine particle changes were smaller, and the children’s respiratory measures did not change.
Fourteen homes is a small sample, and hoods changed along with habits, so the drop cannot be credited to behavior alone. No stove was replaced.
Eight things you can do with the kitchen you already have
- Turn the fan on before the burner, and leave it running for a few minutes after you finish.
- Cook on the back burners whenever the dish allows. This is the change with the clearest measured effect and the lowest cost.
- Find out whether your hood vents outdoors or recirculates. If there is no duct, it recirculates.
- If it recirculates, open a window while you cook and run the fan anyway for grease capture.
- Put a lid on the pan. What stays in the pot does not end up in the room.
- Match the pan to the burner so flames do not run up the sides, and keep burners clean so they burn evenly.
- Wash the metal mesh grease filter in the hood. A clogged filter chokes the airflow you are counting on.
- Test the carbon monoxide alarm this month and check how old it is.
If the fan is too loud to tolerate on high, use it on low every single time rather than on high once in a while. Consistency beats intensity, because the pollutants show up every time you cook.
What this adds up to
Nobody should read this as instruction to replace an appliance. GAO notes that switching stove types after a home is built is uncommon, because running a gas line or upgrading to 240-volt electric service is expensive. Renters do not get a vote at all. An article that ends by telling you to buy a different stove has failed most of its readers.
The more honest summary is that the appliance is the part you probably cannot change and the air is the part you can. On that point the disagreement mostly disappears. The researchers most concerned about gas stove emissions and the reviewers who think the epidemiology is too weak to prove causation both agree that emissions vented outdoors affect nobody.
The advice is also fuel-neutral, which is the detail the public argument lost. If you cook on induction and sear a steak twice a week without turning on the hood, you have a particle problem no stove swap would have solved.
Two situations deserve a clinician rather than a kitchen fix. If a child in the house has asthma that is not well controlled, that belongs in a conversation with their doctor, and the kitchen changes are worth raising there rather than used in place of it. And if anyone gets headaches, nausea or confusion that improve away from home and return when they come back, say so at the appointment. Clinicians are trained to recognize that pattern.
Then go turn on the fan.
Sources
- U.S. Government Accountability Office. “Gas Stoves: Risks and Safety Standards Related to Products and Ventilation (GAO-25-107514).” gao.gov
- U.S. Government Accountability Office. “Gas Stoves: Risks and Safety Standards Related to Products and Ventilation (GAO-25-107514), full report PDF.” gao.gov
- U.S. Environmental Protection Agency. “Carbon Monoxide's Impact on Indoor Air Quality.” epa.gov
- Centers for Disease Control and Prevention. “About Carbon Monoxide.” cdc.gov
- U.S. Environmental Protection Agency. “Indoor Particulate Matter.” epa.gov
- U.S. Environmental Protection Agency. “Improving Indoor Air Quality.” epa.gov
- PubMed. “Kashtan Y, et al. "Nitrogen dioxide exposure, health outcomes, and associated demographic disparities due to gas and propane combustion by U.S. stoves." Science Advances, 2024. DOI: 10.1126/sciadv.adm8680.” doi.org
- PubMed. “Kashtan YS, et al. "Gas and Propane Combustion from Stoves Emits Benzene and Increases Indoor Air Pollution." Environmental Science & Technology, 2023. DOI: 10.1021/acs.est.2c09289.” doi.org
- PubMed. “Lin W, Brunekreef B, Gehring U. "Meta-analysis of the effects of indoor nitrogen dioxide and gas cooking on asthma and wheeze in children." International Journal of Epidemiology, 2013. DOI: 10.1093/ije/dyt150.” doi.org
- PubMed. “Li W, et al. "Gas cooking and respiratory outcomes in children: A systematic review." Global Epidemiology, 2023. DOI: 10.1016/j.gloepi.2023.100107.” doi.org
- PubMed. “Wallace L, Wang F, Howard-Reed C, Persily A. "Contribution of gas and electric stoves to residential ultrafine particle concentrations between 2 and 64 nm." Environmental Science & Technology, 2008. DOI: 10.1021/es801402v.” doi.org
- PubMed. “Holm SM, et al. "Measured air quality impacts after teaching parents about cooking ventilation with a video: a pilot study." Journal of Exposure Science & Environmental Epidemiology, 2024. DOI: 10.1038/s41370-024-00730-6.” doi.org
- PubMed. “Sun L, Wallace LA. "Residential cooking and use of kitchen ventilation: The impact on exposure." Journal of the Air & Waste Management Association, 2021. DOI: 10.1080/10962247.2020.1823525.” doi.org
- PubMed. “Gabdrashova R, et al. "The impact on heart rate and blood pressure following exposure to ultrafine particles from cooking using an electric stove." Science of the Total Environment, 2020. DOI: 10.1016/j.scitotenv.2020.141334.” doi.org