No, Electric Cars Don't Catch Fire More Often Than Gas Cars — The Data Says the Opposite
The video starts with a wail of sirens. A sedan sits in the middle of an intersection, orange flames licking from beneath its doors, a column of black smoke rising into the sky. The caption, shared hundreds of thousands of times, reads: “This is what happens with EVs. They just burst into flames.” The comments fill up fast. “Never buying electric.” “My gas car will never do that.” “Government pushing death traps.”
Scenes like this one have become a fixture of social media. A battery fire is dramatic, unfamiliar, and terrifying — and every one of them seems to be filmed, posted, and shared within hours. The impression left behind is powerful and simple: electric cars catch fire constantly, and gasoline cars do not.
But the impression is wrong. Across nearly every dataset that researchers, insurers, and fire agencies have produced — in the United States, Sweden, Australia, and Poland — the same pattern holds: electric vehicles catch fire at a fraction of the rate of gasoline cars. Gasoline cars, in fact, burn so often that fire authorities count their fires in minutes between incidents, not days. The claim that EVs catch fire more often than gasoline cars is one of the most thoroughly documented myths in the energy-transition debate. The numbers do not just fail to support it. They point in the opposite direction.
That does not mean EV fires are trivial. When a lithium-ion battery pack does go into thermal runaway, the resulting fire is hotter, longer-lasting, and harder to put out than a conventional fuel fire, and it can reignite hours after it appears to be extinguished. Fire departments have had to invent new tactics, new equipment, and new training to deal with it. Honest discussion of EV safety has to hold both facts at once: these fires are genuinely nasty, and they are genuinely rare. What follows is what the fire-incident data actually says — rates, not raw counts, because rates are the only way to compare a fleet of hundreds of millions of gasoline cars with a much younger, much smaller electric fleet fairly.
The claim, and where it comes from
The claim appears in many forms. Sometimes it is blunt — “EVs catch fire more than gas cars.” Sometimes it is hedged — “I keep seeing EVs on fire, I’m not convinced they’re safe.” Sometimes it is dressed up as a consumer warning: a popular meme shows a crowd pointing at a burning orange car, screaming that it must be electric, even though the license plate reads “PETROL.”
The claim feeds on a real asymmetry of experience. Most people have now seen an electric vehicle on fire — on video, at least. Few have ever seen a gasoline car on fire in person, even though gasoline-car fires are enormously more common. That mismatch between what we see and what is true is the whole engine of this myth, and it runs on two fuels: the visual drama of battery fires, and the near-invisibility of ordinary fuel fires in the news cycle.
Consider the arithmetic of attention. According to data compiled by the U.S. Fire Administration and cited by the National Fire Protection Association (NFPA), America experiences on the order of 170,000-plus highway vehicle fires every year, the overwhelming majority involving gasoline or diesel vehicles. Put another way, NFPA figures cited by the motoring organization AAA place the rate at roughly one gasoline-vehicle fire every two to three minutes, around the clock, all year. Nobody films them. Nobody shares them. A gas-car fire on the highway is so routine that it is not news at all.
An EV fire, by contrast, is still unusual enough to be remarkable, visual enough to be shareable, and unfamiliar enough to feel alarming. Professor Paul A. Kohl of Georgia Tech’s School of Chemical and Biomolecular Engineering put it plainly in a 2023 interview with IEEE Spectrum: the media does not treat EVs and ICE vehicles with equal footing, because gasoline is not sensational anymore. One EV fire can generate more coverage than thousands of gas-car fires combined. Over time, those images accumulate into a cultural impression that has nothing to do with probability.
There is also a second, more innocent version of the error: comparing raw counts instead of rates. The United States has hundreds of millions of gasoline vehicles on the road and far fewer EVs, so of course the raw number of gasoline fires dwarfs the raw number of EV fires — and a person looking at raw totals can be persuaded in either direction depending on the point being made. Safety analysts do not compare totals. They compare standardized rates: fires per 100,000 vehicles, or fires per billion miles driven, so the comparison reflects risk rather than market share. On every one of those measures, the result is consistent.
The American numbers: 25 versus 1,530
The most widely cited American comparison comes from an analysis compiled by AutoInsuranceEZ, which cross-referenced incident data from the National Transportation Safety Board (NTSB), registration data from the Bureau of Transportation Statistics (BTS), and government recall databases to produce fire rates per 100,000 vehicles sold, broken down by powertrain type. The figures:
- Hybrid vehicles: 3,474.5 fires per 100,000 vehicles
- Gasoline vehicles: 1,529.9 fires per 100,000 vehicles
- Battery-electric vehicles: 25.1 fires per 100,000 vehicles
In other words, a battery-electric vehicle in this dataset caught fire roughly 61 times less often than a gasoline vehicle — and it was the hybrid, carrying both a fuel tank and a high-voltage battery, that posted the highest fire rate of all.
The hybrid result surprises many people, but engineers are not shocked by it. A hybrid has two energy systems — a tank of volatile liquid fuel and a high-voltage battery pack — and roughly twice the potential points of failure. It also has an important lesson built into it: the data are not simply flattering whatever is newest. Hybrids are an electrified technology too, and they come out worst. The battery-electric cars, with their single energy system, come out best.
Caveats are warranted, and the honest ones cut both ways. The NTSB/BTS-derived figures measure fires per 100,000 vehicles sold, which is a reasonable exposure proxy but not a perfect one. The EV fleet is also much younger on average than the gasoline fleet, and older cars — with degrading wiring, aging fuel lines, and years of deferred maintenance — are more fire-prone. Critics have argued that this fleet-age difference flatters the EV numbers. That objection is worth taking seriously, but it does not erase a sixtyfold gap. Even generous adjustments for fleet age leave EVs far ahead.
The NFPA itself has struck a careful note. In guidance summarized by AAA, the association has said plainly that U.S. data systems do not yet adequately measure EV fire frequency at scale — EV fire-reporting standards and fire-incident coding are still catching up with the technology — and that there is no current evidence that EVs are more likely to catch fire than internal-combustion vehicles. That is not the same as saying the data are perfect. It is the fire-safety establishment saying that everything it can see points one way, while acknowledging the limits of its instruments.
The Swedish data: twenty to one
If the American figures leave you uneasy — perhaps because they are built from an insurance-industry analysis of government data rather than a single clean study — Sweden offers the cleanest natural experiment available. It is one of the world’s most electrified car markets, and its Civil Contingencies Agency (MSB) tracks vehicle fires nationally.
For the year 2022, the agency’s numbers were stark. Among roughly 611,000 electric vehicles on Swedish roads, there were just 23 fires — a fire incidence of 0.004 percent. In the same period, among approximately 4.4 million gasoline and diesel vehicles, there were 3,400 fires — an incidence of about 0.08 percent. Combustion-engine vehicles were about twenty times more likely to catch fire than EVs, in a country where EVs were already a substantial share of the fleet.
Other cuts of the Swedish data tell the same story. Separate analyses of the country’s figures have put petrol and diesel cars at around 68 to 77 fires per 100,000 registered vehicles, versus roughly 3.8 per 100,000 for battery-electric vehicles — again, about twenty times lower for EVs. Norway’s Directorate for Social Security and Emergency Preparedness has found a similar pattern: four to five times more fires in gasoline and diesel cars compared with EVs. Different countries, different agencies, different methodologies — the same direction.
Australia and Poland: the pattern repeats
Australia’s EV FireSafe project maintains a verified global database of electric-vehicle fire incidents, and its analysts have converted that incident tracking into an estimated EV fire probability of 0.001 to 0.002 percent, compared with roughly 0.1 percent for petrol and diesel vehicles. That places combustion vehicles at a 50- to 100-fold higher fire risk, depending on assumptions about fleet age and reporting completeness.
Poland offers yet another independent check, this time from the State Fire Service, cited by the Polish New Mobility Association in 2025. Between 2020 and 2025, Poland recorded 51,142 vehicle fires. Of those, 50,833 involved internal-combustion vehicles. Hybrids accounted for 222. Electric vehicles accounted for just 87.
The Polish figures are raw counts rather than rates, and as noted above, raw counts flatter whichever fleet is larger. But the Polish EV fleet over that period was not one-eightieth the size of the ICE fleet in fire-count terms — it was far larger than 87 fires’ worth — and even as raw numbers, the contrast is overwhelming. When four independent national datasets, collected by fire services, safety agencies, and verified incident trackers, all point the same way, the shape of the evidence is hard to dismiss.
Per mile: Tesla’s long-running data
A second way to normalize the comparison is by miles driven rather than by vehicles, and here the most detailed long-run dataset comes from a single manufacturer: Tesla, which has published vehicle fire data for its fleet for years. Between 2012 and 2023, Tesla reported approximately one vehicle fire for every 135 million vehicle miles traveled — and the company counts every known fire event in that figure, including fires that did not start with the car at all, such as structure fires, wildfires, and arson, in order to align with how agencies like the NFPA collect their data.
The U.S. national average over comparable periods has been roughly one vehicle fire per 17 to 19 million miles. On Tesla’s own accounting, a Tesla was therefore about eleven times less likely to experience a fire per mile driven than the average American vehicle. A later company report covering 2012 through 2024 put the Tesla figure at about one fire per 140 million miles — the same picture, essentially unchanged.
As with the per-vehicle data, there are fair objections. Tesla’s fleet is young, well-maintained, and disproportionately composed of newer vehicles; the company is also the source of its own statistics, which invites skepticism. But the per-mile methodology is the right one — it corrects for the fact that a parked car and a car driven 30,000 miles a year do not have the same fire exposure — and the gap is again far too large to be explained away by fleet age alone.
The part the myth gets right: EV fires are nastier
Everything above concerns frequency: how often vehicles catch fire. The claim under examination — “electric cars catch fire more often than gasoline cars” — is a claim about frequency, and it is false. But there is a related set of claims, often bundled with it, where the EV skeptics have a genuine point: when an EV does catch fire, the fire behaves differently and more dangerously than a gasoline fire.
A lithium-ion battery fire is a chemical event called thermal runaway. One damaged or defective cell overheats, and the heat cascades from cell to cell, each one releasing its stored energy. Because the battery carries both its fuel and its oxidizer, the reaction can sustain itself without outside oxygen the way a conventional fire requires. That is why EV fires burn hotter, last longer — sometimes hours — and demand enormous quantities of water to cool the pack rather than simply suppress flames. Firefighters have reported using tens of thousands of gallons on a single EV fire; in one widely reported April 2021 incident near Houston, Texas, crews spent about four hours and more than 30,000 gallons of water bringing a Tesla fire under control, something the responders said they would normally have handled in minutes with a gasoline car.
Then there is reignition. A gasoline fire, once its fuel is consumed or cut off, stays out. A damaged battery pack can reignite hours or even days after the visible fire is extinguished, as damaged cells continue their slow cascade — which is why fire departments now monitor and quarantine burned EVs rather than simply towing them away. The runoff from firefighting water can also carry leached battery chemicals, adding an environmental-hazard dimension that gasoline fires largely lack.
An in-depth report produced for the U.S. National Highway Traffic Safety Administration (NHTSA) summarized the state of the science this way: “The propensity and severity of fires and explosions from lithium-ion battery systems are anticipated to be somewhat comparable to or perhaps slightly less than those for gasoline or diesel vehicular fuels,” while cautioning that the technology and industry had not yet matured enough to have established comprehensive safety codes and standards.
That is a carefully worded, genuinely expert verdict, and it is worth sitting with. Propensity and severity — comparable to, or slightly less than, gasoline. Not zero. Not trivial. Not a reason to panic, and not a reason for complacency either. Fire departments are adapting: EV-specific protocols, thermal-imaging monitoring, piercing nozzles, fire blankets, updated training. The adaptation is real, and it is evidence that professionals take battery fires seriously — as a rare but severe hazard, which is exactly what the data says they are.
Why hybrids burn most, and what it teaches us
The AutoInsuranceEZ finding that hybrids have the highest fire rate — about 3,475 per 100,000, more than double the gasoline rate — deserves a closer look, because it demolishes the lazy version of the debate in both directions. If you believed “newer technology is automatically safer,” the hybrid result refutes you. If you believed “anything with a battery is a fire trap,” the battery-electric result refutes you.
The engineering explanation is straightforward. A hybrid carries two complete energy systems: a gasoline fuel system with all its familiar failure modes (leaking lines, hot exhaust components, electrical shorts in an aging 12-volt system) plus a high-voltage traction battery and its management electronics. More systems, more complexity, more failure points. A battery-electric vehicle, by contrast, has one energy system and far fewer moving parts, no hot exhaust, no fuel pump, no fuel lines snaking past a hot engine. Gasoline is an energy-dense liquid that is specifically designed to burn; an EV battery is an energy-dense solid that is specifically designed not to.
This also helps explain a common confusion in the raw data. In many fire-incident databases, hybrid fires get coded in ways that blur the lines — a hybrid burning because of a fuel-system fault can be tallied in a way that muddies the “EV” category in public discussion. When researchers separate battery-electrics from hybrids, as the NTSB/BTS analysis does, the battery-electrics come out clearly lowest.
Why the myth persists
By now the shape of the evidence is clear, so it is worth asking why the claim remains so sticky. The research on risk perception offers several answers, and they all apply here.
First, the availability heuristic: people judge frequency by how easily examples come to mind. EV fires are filmed, shared, and replayed; gasoline fires are not. The example that comes to mind is the EV fire, so the EV fire feels common. The estimated 170,000-plus American highway vehicle fires each year, nearly all of them gasoline, are invisible — they happen one every couple of minutes, unrecorded, unshared, unremarked.
Second, novelty and dread. Psychologists have long found that unfamiliar, hard-to-control hazards feel scarier than familiar ones, even when the familiar ones kill more people. Everyone understands a gas-tank fire; almost nobody intuitively understands thermal runaway. The unfamiliar hazard gets more mental weight.
Third, misattribution. Viral footage of burning vehicles circulates without context, and sometimes a burning car is simply declared an EV when it is not. A widely shared meme shows an outraged crowd insisting a burning orange car must be electric while its license plate reads “PETROL.” The joke works because it is true to life: the assumption now runs ahead of the evidence.
Fourth, motivated reasoning. The EV transition is politically charged, and vehicle fires have become ammunition in a culture war that has little to do with fire statistics. A dramatic video is a better weapon than a spreadsheet, and nobody’s mind was ever changed by a rate per 100,000.
None of this means every person who worries about EV fires is irrational. The fires are real, the footage is real, and the difficulty of extinguishing them is real. But worry calibrated to frequency — the actual question the claim asks — points the other way.
The honest caveats
A fair fact-check should steelman the other side, and there are three caveats that deserve to be stated plainly.
One: the EV fleet is young. The average electric vehicle on the road is much newer than the average gasoline vehicle, and vehicle age correlates with fire risk — wiring degrades, fuel systems corrode, maintenance lapses. As today’s EVs age into their teens and twenties, their fire rates will almost certainly rise from current levels. The question is whether they will rise to anywhere near gasoline levels, and nothing in the current data suggests they will, given the sixtyfold starting gap.
Two: reporting is imperfect. The NFPA itself has acknowledged that American data systems do not yet adequately capture EV fire frequency at scale. Definitions vary, incident coding is inconsistent, and some EV fires may be misclassified. This cuts both ways — imperfect data could be hiding EV fires or overstating the gap — but the direction of every independent dataset is the same, and “the data are imperfect” is not evidence that the conclusion is wrong.
Three: frequency is not the whole of safety. Even if EVs catch fire far less often, a single severe battery fire in an underground garage or on a car carrier raises legitimate questions about consequences, not just probabilities. Fire services, building codes, and ferry operators are right to plan for the severe-but-rare event. Conflating “rare” with “nothing to worry about” is its own kind of error.
These caveats qualify the conclusion. They do not overturn it.
The Verdict
The claim that electric cars catch fire more often than gasoline cars is False.
Across the best available fire-incident data — NTSB and BTS figures compiled for the U.S. market (25.1 EV fires per 100,000 vehicles versus 1,529.9 for gasoline, roughly a 61-fold difference), Swedish national fire data (combustion vehicles about 20 times more likely to burn than EVs), Australia’s verified EV fire-incident tracking (a 50- to 100-fold gap), Polish fire-service records (87 EV fires out of 51,142 vehicle fires over five years), and Tesla’s per-mile fleet data (about one fire per 135 million miles versus roughly one per 17 to 19 million nationally) — electric vehicles catch fire dramatically less often than gasoline cars. Hybrids, carrying both a fuel tank and a battery, actually post the highest fire rates of all.
What is true, and what the myth distorts, is that EV fires are harder to extinguish, burn longer, and can reignite — a genuine hazard that fire departments are rightly adapting to. Rarity and severity are different things. The data says EV fires are rare. The chemistry says they are severe. Both statements are true at the same time, and any honest conversation about electric-vehicle safety has to include both.
So the next time a video of a burning EV races across your feed, remember the arithmetic of attention. You are watching one of the rarest kinds of car fire on Earth — which is precisely why someone filmed it. The common kind, the gasoline kind, is happening somewhere right now, roughly every two to three minutes, and nobody is filming that. The camera does not measure risk. The data does — and the data says the electric car in your driveway is far less likely to catch fire than the gasoline car it replaced.