Anything that burns a fuel inside your house produces gases that shouldn’t end up in the air you breathe. That’s the whole topic in one sentence. The interesting part is that the risks sort cleanly into three tiers, and each tier deserves a different kind of attention.

The three tiers, plainly:

  1. Carbon monoxide is the acute risk. Rare, but when it happens it kills quickly. The fix is cheap and well-understood: alarms.
  2. Nitrogen dioxide from gas cooking is the chronic risk, especially for kids’ developing lungs. The fix is mostly about ventilation — running a range hood that actually goes outside.
  3. The broader gas-stove debate is somewhere in between — real underlying physics, mixed evidence on broader health effects, policy questions that go well past pure science.

This article walks through each tier with the same building-science voice the rest of the site uses for mold: describe what’s well-established, point at what’s contested, and route to fixes that help regardless of which way the contested parts eventually break.

Carbon monoxide — the acute risk

CO is a colorless, odorless, tasteless gas produced whenever a carbon-based fuel — natural gas, propane, oil, gasoline, wood, charcoal, kerosene — burns incompletely. It binds to hemoglobin roughly 200 to 250 times more strongly than oxygen, so even modest concentrations starve tissues of the oxygen they need. There is no human sense that detects CO before symptoms begin. That’s why detectors aren’t optional.

The U.S. averages somewhere over 400 accidental CO deaths per year and tens of thousands of emergency-room visits, with the toll concentrated in winter (heating equipment plus closed-up houses) and after major power outages (generators). The CDC’s CO information hub is the canonical reference; we’ll summarize the parts that change how you live in your house.

Where it comes from

The recurring list of household CO sources is short and worth knowing:

  • A furnace or boiler with a cracked heat exchanger, a blocked flue, or a chimney that’s started to backdraft.
  • A gas water heater with an atmospheric (natural-draft) vent that spills exhaust when the house is depressurized — the classic trigger is a strong kitchen range hood plus a bath fan plus a dryer all pulling air out at once on a tight house.
  • A fireplace with a closed or partially-closed damper, or a chimney with a bird’s nest in it.
  • A wood stove or pellet stove with poor draft, blocked flue, or creosote buildup.
  • A gas range or oven used for space heating — a major poisoning pattern in cold-weather, low-income households.
  • A portable generator run in a garage, basement, or near an open window after a storm. This is the single most-deadly CO scenario after hurricanes and ice storms.
  • A car idling in an attached garage. Modern keyless-ignition cars make this worse — people walk away thinking the engine is off when it isn’t.
  • A charcoal grill, camp stove, or kerosene heater brought inside. Never. Not for cooking, not for warmth, not “just for a few minutes.”

Symptoms — and why they get missed

Headache, dizziness, weakness, nausea, confusion, chest pain. The symptom set overlaps almost perfectly with the flu, and with several other indoor-environmental complaints. The classic tells:

  • Symptoms that improve when you leave the house and return when you go back in.
  • More than one person in the household feeling sick at the same time, including pets.
  • Symptoms that cluster around heating season or around the use of a specific appliance.

The worst poisonings happen at night, because sleep masks the early symptoms long enough for the exposure to reach lethal levels. That’s the entire reason the alarm recommendation is what it is.

Detection — the one piece of safety gear you cannot skip

The U.S. CDC, NFPA 72, and essentially every state code converge on the same residential CO-alarm placement:

  • One alarm on every level of the home, including the basement.
  • One within 10 feet of every sleeping area — or, the modern upgrade, one inside each bedroom for households with heavy sleepers or hard-of-hearing occupants.
  • UL-2034 certified units. This is the U.S. residential CO-alarm standard. Anything sold legitimately as a “carbon monoxide alarm” will say UL 2034 somewhere on the back.
  • Not mounted right next to fuel-burning appliances (false positives), in the garage (temperature range), or in damp areas.

A $25 alarm with a 10-year sealed battery, on each level, near the bedrooms, is the floor. The math on this is unambiguous — you cannot spend money on home safety equipment more efficiently than this.

Replacement cycle: five to ten years depending on the unit. The sensor degrades. A “working” twelve-year-old alarm that still chirps on its test button may not actually detect CO anymore. Modern units have a printed install date and an end-of-life chirp; the simple habit is write the install date on the unit in Sharpie when you put it up, and replace per the manufacturer’s stamped life.

Combination smoke / CO units are fine and code-compliant in most jurisdictions. The tradeoff is that when either sensor expires the whole unit goes — slightly worse economics, slightly better compliance because there are fewer devices to mount.

Low-level monitors (devices like the NSI 3000, CO Experts, and some Forensics Detectors models) read down to single-digit ppm and alarm in ranges UL 2034 deliberately ignores. These are a worthwhile upgrade for households with infants, pregnant occupants, cardiac patients, or anyone who’s read enough about the HVAC trade to want more than the floor. They are not a replacement for a UL-2034 alarm; they are a layer on top.

What to do if a CO alarm goes off

The CDC / CPSC / fire-service consensus is short:

  1. Get everyone out. Into outside air, immediately. Don’t open windows first. Don’t investigate. Leave.
  2. Call 911, or the local fire department’s non-emergency line if no one is symptomatic. Fire crews carry calibrated CO meters and will sweep the house.
  3. Don’t go back in until the source is identified and the levels verified safe.
  4. Seek medical attention if anyone is symptomatic. A carboxyhemoglobin blood test is the standard.
  5. Find and fix the source before you re-fire whatever appliance was likely responsible. HVAC service, chimney sweep, gas utility — most utilities will send a tech for a suspected leak at no cost, day or night.

The framing matters. Assume the alarm is real; verify after. The opposite habit — “it’s probably nothing, the alarm is being annoying” — is how people die.

Prevention

Mostly maintenance, mostly boring, mostly cheap:

  • Annual servicing of every fuel-burning appliance — furnace, boiler, water heater, fireplace. A combustion analyzer in the hands of a competent tech catches problems before they catch you.
  • Annual chimney inspection for any wood, pellet, or gas fireplace, and any vented stove.
  • Never run a generator indoors, in an attached garage, or within 20 feet of a window or door — ever, under any conditions. Most post-storm CO deaths are this single error.
  • Never warm up a car in an attached garage, even with the door open. Back it out first.
  • Never bring a charcoal grill, camp stove, or unvented combustion heater into the house. They are outdoor equipment.

Nitrogen dioxide — the chronic risk from gas cooking

Where CO is about acute poisoning, NO2 is about chronic respiratory irritation. It’s the byproduct that puts gas stoves in the public-health conversation.

What’s actually happening

When a gas (or propane) burner runs, the flame is hot enough to oxidize nitrogen from the air, producing nitric oxide (NO) and nitrogen dioxide (NO2). NO2 is the one with the well-established respiratory profile — it’s a known irritant, regulated outdoors as a criteria pollutant, and linked in the medical literature to asthma exacerbation. Electric resistance and induction cooktops, by contrast, don’t burn anything. They don’t produce NO2 at all.

EPA’s outdoor 1-hour standard for NO2 is 100 ppb. Indoor air doesn’t have to meet that standard — there’s no federal indoor NO2 limit — but peer-reviewed measurements in small, poorly-ventilated kitchens routinely cross 100 ppb within minutes of turning on a gas burner. If the EPA outdoor limit did apply indoors, a lot of American kitchens would be in violation during dinner.

The EPA’s overview of indoor nitrogen dioxide is the clearest one-page summary.

What the research says, fairly

The most-cited recent work: a 2022 population-attributable-fraction analysis (Gruenwald et al.) estimated that around 12% of current childhood asthma in the U.S. is attributable to gas-stove use, drawing on an earlier meta-analysis (Lin et al. 2013) that found a pooled odds ratio around 1.32 — so roughly a 30% elevation in childhood asthma odds for children in gas-cooking households. That estimate sits on top of decades of indoor-air work showing NO2 from gas cooking is a real and measurable exposure.

Fair caveats, named the way we name them everywhere else on this site:

  • A population attributable fraction is a what-if estimate, not a measured causal effect.
  • The underlying odds ratios come largely from cross-sectional studies where reverse causality and confounding can’t be fully ruled out.
  • Newer cohort work has sometimes found weaker or null associations.
  • The biological plausibility, though, is genuinely solid. NO2 is a characterized respiratory irritant with regulatory exposure limits; measured indoor concentrations during gas cooking are not trivial; children’s airways are more vulnerable than adults’.

Calibrating the magnitude: this is real, not catastrophic. Kids in households with chronically high NO2 exposure and other asthma risk factors are most affected. Adults with existing asthma may notice symptoms during or after cooking. Healthy adults probably don’t notice in any measurable way. The honest summary is that gas cooking is one modifiable input among several for childhood asthma — worth taking seriously, not worth panicking about.

The single best mitigation

Vent your range hood outside, every time you cook, on a speed appropriate to the heat.

That’s the whole intervention, and it’s the one with the most leverage. Three things have to be true for it to work:

  1. The hood actually vents outside. Open the cabinet above the hood, follow the duct, and confirm it terminates at a wall- or roof-cap on the exterior of the house. If the duct ends inside the cabinet — or if there’s no duct at all and the hood just blows back into the kitchen through a charcoal filter — your hood is recirculating, and it is not helping with NO2. Recirculating hoods catch grease and odor. They do almost nothing for combustion gases.
  2. You actually use it. Every burner, every cook. Not just when there’s smoke.
  3. You run it on the back burners when you can. Hood capture efficiency is much higher over the back of the cooktop than the front. This is a free improvement.

A useful add: leave the hood on for five to ten minutes after you finish cooking. The combustion plume keeps coming off hot pans and the cooktop for a while after the flame is off.

Other mitigations

In rough order of impact:

  • Crack a window while cooking, especially if your hood is weak, recirculating, or missing. Better than nothing by a lot.
  • Cover pots with lids when possible — less cook time, less combustion, less moisture in the air (a moisture-control bonus too).
  • When the stove eventually needs replacing, consider induction. Faster than gas, cleaner than gas, no combustion byproducts, and the cookware compatibility issue is mostly solved at this point. We’re not telling anyone to rip out a working appliance; we’re saying that when replacement comes due, induction is worth a real look for IAQ, speed, and energy reasons together.

The broader politicized debate

This is the part of the topic where the political heat exceeds the scientific heat, and where readers come in with strong priors in both directions. The calm version:

What’s well-established:

  • Gas combustion produces NO2.
  • Indoor NO2 in poorly-ventilated kitchens routinely exceeds the EPA outdoor standard during cooking.
  • NO2 is a respiratory irritant with documented effects on asthma symptoms.
  • The childhood-asthma association with gas cooking has held up in meta-analyses, with the caveats above.

What’s contested or less clear:

  • The size of the effect in any given household — small kitchen vs. large, hood vs. no hood, frequency of cooking, ventilation rates, vulnerable occupants vs. not.
  • Broader health claims about adult illness from gas cooking — the evidence here is thinner than the childhood-asthma evidence.
  • The methane-leakage angle — gas stoves leak small amounts of unburned natural gas even when off. The climate implication is real; the indoor-air implication for typical homes is much less clear.

What was mostly misreporting:

  • The 2023 “gas-stove ban” story. What actually happened: a single CPSC commissioner said in an interview that regulatory options were on the table. The framing became “the federal government is going to ban gas stoves.” CPSC and the White House clarified within days that there was no proposed ban. There still isn’t. There are state and city policies — new-construction gas-hookup bans in some cities, pre-emption laws preventing such bans in other states — but at the federal residential-appliance level, the “ban” was never the story it was reported as.

The honest version of all of this: gas cooking has real measurable air-quality effects in the kitchen; range-hood ventilation matters a lot; electrification is one valid policy direction among several; plenty of households cook on gas safely with proper ventilation.

We try to avoid both the wellness-influencer framing (“your gas stove is poisoning your family”) and the dismissive framing (“this is just political theater”). Both miss the actual building-science point: U.S. residential ventilation has not kept up with how we cook. The fix is boring and apolitical — make the hood vent outside, use it every time.

Other combustion sources — wood and the romance problem

Solid-fuel appliances are where the photogenic and the air-quality inconvenient meet.

Wood stoves. Even when working well, they emit meaningful PM2.5 (fine particulate). EPA-certified modern units emit a small fraction of what 1980s and 1990s stoves emit; if you heat with wood and have an old non-certified stove, replacing it is one of the higher indoor-and-outdoor air-quality upgrades available to you. Burn dry seasoned wood (≤20% moisture), build hot fires rather than smoldering ones, and get the chimney swept annually.

Open masonry fireplaces. From an air-quality and efficiency standpoint, these are the worst combustion appliance in the house. They exhaust most of their heat up the flue, depressurize the room pulling outside air in (and potentially backdrafting other appliances), and routinely leak smoke into the room on startup and under cold-flue conditions. They are best understood as ambiance, not heat. We’re not telling you not to enjoy your fireplace. We’re telling you to be honest with yourself about what it’s doing.

Pellet stoves. Generally the cleanest of the solid-fuel options for in-home use — forced draft, standardized fuel, much lower PM than either open fireplaces or older wood stoves. Still meaningful particulate; still want the room CO alarm and the annual chimney work.

The cooking-and-moisture overlap

Worth pointing out because it lands in both this article and the moisture-control guide. A four-person household generates several gallons of water vapor per day from cooking, bathing, breathing, and indoor laundry drying. The range hood that handles your NO2 problem is the same range hood that handles a meaningful chunk of your kitchen-moisture problem. One piece of equipment, two jobs, the same use pattern: vent outside, run it during cooking, run it for a few minutes after.

If you’re already running the hood for moisture-control reasons, the NO2 mitigation is a free side effect. If you’re already running it for NO2 reasons, the moisture mitigation is the bonus. The single habit that gets both is use the hood, every cook.

HVAC, water heaters, and backdrafting

A small but important corner of this topic that doesn’t get enough attention in consumer coverage.

Sealed-combustion (“direct-vent”) appliances are the modern standard for new installs. A direct-vent furnace or water heater pulls its combustion air from outside through one pipe and exhausts through another, completely isolated from the indoor air. The flame might as well be in a sealed box on the outside wall. These are inherently safe against backdrafting and are the right thing to install when an appliance is up for replacement.

Atmospheric-vent appliances are the older standard — a burner with an open hood at the top, exhausting up through a metal flue into a chimney by natural draft. They’ve worked fine for decades, but they have one quiet failure mode: under low indoor-pressure conditions, the flue can become the makeup-air path for the house, pulling combustion gases back into the room instead of out of it. The classic trigger is a tight modern house with a powerful range hood running while the furnace and water heater are on at the same time. Telltales: rust streaks or soot above the draft hood, condensation on the flue, a sweet/sharp combustion smell near the appliance.

If you have atmospheric-vent appliances and a tight house with strong exhaust fans, two reasonable moves: (1) when an appliance comes due for replacement, replace it with a sealed-combustion unit; (2) in the meantime, make sure your CO alarms are placed and current. This is one of the scenarios they are most useful for.

Testing for CO and NO2

A short reality check on what’s worth measuring at home.

Carbon monoxide. $25–50 alarms, ubiquitous, no excuse not to have them. Replace per the unit’s stamped life. For households with vulnerable occupants or anyone who wants quantitative readouts, consumer-grade low-level monitors run $150–300 and are a reasonable upgrade — they give you a number instead of just an alarm threshold. Either way, this is a solved problem at the consumer level.

Nitrogen dioxide. Harder. Some consumer indoor-air monitors (certain Airthings models, Awair, a handful of others) include NO2 sensors, and accuracy varies considerably between models and over time. Useful as awareness tools — “wow, my NO2 spikes when I cook” is a real and motivating observation — but generally less actionable than just running the hood. If the action is the same either way, the measurement is optional.

Multi-parameter indoor-air-quality monitors that report PM2.5, VOCs, CO2, humidity, and (sometimes) NO2 are useful for awareness and diagnosing patterns. They are not safety devices. Don’t confuse them with CO alarms; they don’t substitute. We cover monitor selection in more detail elsewhere on the site as those pages land.

What to do today

If you read nothing else above, this is the short list:

  1. Walk the house and count your CO alarms. One on every level, one near every sleeping area. UL-2034 certified. Check the stamped replacement date on each. Replace anything older than its stamp, anything without a stamp, or anything you can’t find at all.
  2. Write the install date on each new alarm in Sharpie so the next person to check doesn’t have to guess.
  3. Open the cabinet above your range hood and follow the duct. If it terminates in the cabinet or feeds back into the kitchen through a filter, your hood is recirculating. Note that as a real-but-not- urgent improvement for the next remodel; in the meantime, crack a window when cooking.
  4. If your hood does vent outside, start actually using it — every burner, every cook, on the back burners when you can, for five to ten minutes after.
  5. Schedule annual service for the furnace, boiler, water heater, and any fireplace or chimney.
  6. Never run a generator, grill, or unvented combustion heater indoors or in an attached garage. If a storm is coming and you own a generator, walk now and plan where it will live outside — not in the moment, in the dark, with the power already out.
  7. If you have a working gas stove and you cook with kids at home, the hood habit is the single biggest lever you have. It costs nothing and it works.

The framing we keep coming back to: combustion gases are a venting problem first, a health problem second. The science on the acute risk (CO) is settled and the fix is cheap. The science on the chronic risk (NO2) is real and the fix is mostly ventilation we should already be doing. The broader political debate doesn’t change either of those fixes. You don’t need to wait for the debate to settle to act on the parts that are clear.

That’s the calm version. We’ll take it.