What are the odds of an electric car catching fire?
Evidence quality 4.5/5
Eight-dimension review score against the quality rubric . Each dimension scored 1–5.
- D1 Source grounding
- 4/5
- D2 Source authority
- 4/5
- D3 Arithmetic
- 5/5
- D4 Uncertainty
- 4/5
- D5 Scope
- 5/5
- D6 Prose
- 5/5
- D7 Perception honesty
- 4/5
- D8 Caveat completeness
- 5/5
● your factors — click this risk ▾ to reveal
- Your factors
≈ As likely as
Perceived
Every EV fire generates outsized media coverage. A single Tesla fire on a highway attracts the kind of attention that hundreds of daily gasoline-car fires do not, because EV fires are novel, visually dramatic (thermal runaway produces sustained flames and toxic fumes), and difficult for fire services to extinguish with conventional methods. The result is a perception gap: surveys consistently show that consumers rank battery fire as a top concern about EV ownership, even though the data shows EVs catch fire at a fraction of the rate of internal combustion engine vehicles.
Rough estimate: Many consumers believe EVs are more likely to catch fire than gasoline cars
Source: editorial intuition, not polled
Actual
~25 fires per 100,000 EVs sold vs ~1,530 per 100,000 ICE vehicles (AutoInsuranceEZ/BLS data)
US registered EVs
Show derivation
Uses the widely cited AutoInsuranceEZ/BLS-derived figure of ~25 fires per 100,000 EVs sold, compared to ~1,530 per 100,000 for ICE vehicles. The EV figure is based on NTSB and NHTSA incident data through 2022. Swedish Civil Contingencies Agency (MSB) data confirms the direction: 3.8 fires per 100,000 EVs and hybrids vs 68 per 100,000 ICE vehicles in Sweden, making ICE vehicles about 18x more likely to catch fire. The normalised figure uses the 25/100,000 annual rate (0.00025/year) compounded over a ~12-year average vehicle ownership period: 1 − (1 − 0.00025)^12 ≈ 0.003, yielding ~1 in 333 chance that a given EV will experience a fire at some point during ownership. For ICE vehicles, the equivalent figure is ~1 in 5.5 over the same period. The comparison is the point: EVs are roughly 60x less likely to catch fire than ICE vehicles per the US data, and ~18x less likely per the Swedish data. The uncertainty band reflects the difference between US and Swedish estimates and the fact that the EV fleet is younger than the ICE fleet (older vehicles are more fire-prone).
Caveats: The EV fire rate comparison to ICE vehicles is directionally robust (EVs catch f…
The EV fire rate comparison to ICE vehicles is directionally robust (EVs catch fire far less often) but the magnitude of the gap is uncertain for several reasons. First, the EV fleet is younger on average than the ICE fleet, and vehicle fire risk increases with age — as the EV fleet ages, the rate may converge somewhat. Second, the absolute number of EV fires is small enough that the per-100k rate has wide confidence intervals. Third, EV fires have different characteristics: thermal runaway in lithium-ion batteries produces sustained, high-temperature fires that are harder to extinguish, can reignite days after the initial event, and produce toxic hydrogen fluoride gas. The per-incident severity may be higher for EV fires even if the per-vehicle frequency is lower. Finally, the comparison includes all ICE vehicle fires, many of which are in very old vehicles; comparing EVs to age-matched ICE vehicles would narrow the gap. The entry is tagged as overrated because the data clearly shows EVs are less fire-prone than ICE vehicles, contrary to the popular perception. A separate dimension worth naming is what happens after a fire starts, which is independent of how often one starts. The NTSB's responder-safety report (NTSB/SR-20/01, 2020) found that all three crash-damaged EV batteries it investigated reignited after firefighters had extinguished the vehicle fire — one reignited at least six times, including 5 days after the crash — because of "stranded energy": charge trapped in undamaged cells with no path to discharge, which responders cannot measure or drain except by letting the pack burn itself out. The Australian fire-service project EV FireSafe reports that 13% of the EVs in its reignition-incident dataset reignited after initial suppression, with cases of four separate reignition events, a 68-day delayed reignition, and reignitions occurring while the vehicle was on a tow truck. These reignition figures are conditional on a fire having already occurred (a per-incident severity characteristic), are drawn from small, self-selected fire-service datasets with no population denominator, and do not change the entry's headline ignition odds; they are the reason an EV fire is operationally harder to make safe than an ICE fire of equivalent size, not evidence that EVs ignite more often. The enclosed-structure severity concern (garages, tunnels) is the same severity-not-frequency point and is already reflected in the indoor-charging multiplier above.
Related risks
Other risks on similar themes — for exploring related fears.
Hand-held phone call + driving
What are the odds of a crash from holding a phone to your ear while driving?
Child without restraint
How much more likely is a child to die in a car crash without an appropriate child restraint?
Pick challenger
Electric vehicles catch fire at a rate of roughly 25 per 100,000 vehicles sold, compared to approximately 1,530 per 100,000 for gasoline-powered cars — a ratio of about 60 to 1 in favour of EVs. Swedish data from the Civil Contingencies Agency tells the same story at a different scale: 3.8 fires per 100,000 for EVs and hybrids versus 68 per 100,000 for ICE vehicles, with only 23 fires recorded among 611,000 registered electric cars in the country. The National Fire Protection Association tracks roughly 174,000 highway vehicle fires per year in the United States; EVs are underrepresented in that count relative to their share of the fleet. By every available dataset, electric cars are substantially less likely to catch fire than their gasoline counterparts.
The perception runs in the opposite direction because EV fires are novel, photogenic, and genuinely different in character. Thermal runaway in a lithium-ion battery pack produces a sustained, high-temperature fire that conventional water-based suppression cannot easily extinguish. The fire can reignite hours or days after the initial event. It produces toxic hydrogen fluoride gas. Fire departments have had to develop entirely new protocols for EV fires, including submerging entire vehicles in water containers. These operational challenges are real and worth taking seriously, but they describe per-incident severity, not per-vehicle frequency. The media covers EV fires the way it covers shark attacks: each incident is reported as if it reveals a systemic danger, while the vastly more common gasoline-car fires go unmentioned because they are too routine to be news.
The main caveat is fleet age. The EV fleet is young — most EVs on US roads are less than five years old — and vehicle fire risk increases with age for all powertrain types as wiring degrades, seals fail, and batteries accumulate charge-discharge cycles. As the first generation of mass- market EVs ages past 10 years, the fire rate may increase. Whether it converges toward the ICE rate or remains substantially below it is an open empirical question. Post-collision battery damage is a specific EV risk that the NTSB has flagged: a crash that would leave an ICE vehicle safely disabled can damage a battery pack in ways that trigger delayed thermal runaway hours later, requiring extended post-crash monitoring. The comparison is imperfect, but the direction is clear: the fear that EVs are fire-prone is not supported by the data.
Related tidbits
EV battery fires occur at roughly 25 per 100,000 vehicles. Gasoline car fires: ~1,500 per 100,000. Gas cars catch fire about 60x more often, but "EV catches fire" is a headline. "Gas car catches fire" is not.
Claim ledger
Every number below is what each source reported, with the verbatim quote we relied on and how we arrived at our figure. Click any link to verify directly.
-
[1] Kelley Blue Book / AutoInsuranceEZ — Report: EVs Less Likely to Catch Fire Than Gas-Powered Cars
Report: EVs Less Likely to Catch Fire Than Gas-Powered Cars- Statistic
~25 fires per 100,000 EVs sold vs ~1,530 fires per 100,000 ICE vehicles vs ~3,475 per 100,000 hybrids- Excerpt
“"There are about 25 fires per 100,000 electric vehicles sold compared to 1,530 fires per 100,000 internal combustion engine vehicles sold." ”
- Source data from
- 2023-01-20
- Accessed
- 2026-04-18 · archived copy
- Calculation
- AutoInsuranceEZ compiled data from the Bureau of Transportation Statistics, NHTSA, and NTSB to calculate fire rates per 100,000 vehicles sold by powertrain type. The EV rate of ~25/100k vs ICE rate of ~1,530/100k represents a roughly 60x lower fire rate for EVs. The hybrid rate (~3,475/100k) is highest, likely reflecting the dual powertrain and the older average age of the hybrid fleet. KBB is used as the citation because it is the most widely read consumer source reporting this data. The underlying BLS/NHTSA data is the authoritative source. The figure has been widely cited but carries caveats: the EV fleet is younger on average than the ICE fleet (vehicle fire risk increases with age), and the total number of EV fires is small enough that the per-100k rate has a wide confidence interval.
-
[2] Allied World Insurance — Electric Vehicle Fires: A cause for concern?
Electric Vehicle Fires: A cause for concern?- Statistic
Swedish MSB data: 3.8 fires per 100,000 EVs/hybrids vs 68 per 100,000 ICE vehicles; only 23 fires among 611,000 EVs in Sweden- Excerpt
“"In Sweden, EVs and hybrids caught fire at a rate of 3.8 fires per 100,000 vehicles, whereas combustion-engine cars caught fire at 68 per 100,000." ”
- Source data from
- 2024-03-01
- Accessed
- 2026-04-18 · archived copy
- Calculation
- The Swedish Civil Contingencies Agency (MSB) data is the most frequently cited European dataset on EV fires. Sweden has high EV adoption (~17% of new car sales by 2023) and comprehensive fire reporting. The 3.8 vs 68 per 100,000 comparison shows ICE vehicles are ~18x more likely to catch fire — directionally consistent with the US data but with a smaller ratio, likely because the Swedish ICE fleet is newer on average and Sweden's fire statistics may capture different incident types. The absolute EV fire count (23 fires among 611,000 EVs) illustrates how small the numerator is, which is why the per-100k rate has wide confidence intervals.
- Independence
- Swedish MSB data is collected independently from US NHTSA/NTSB data. The two datasets use different methodologies and cover different vehicle fleets, making them a genuine cross-validation.
-
[3] Fire Rover — How Often Do Electric Cars Catch Fire? A Look at the Statistics
How Often Do Electric Cars Catch Fire? A Look at the Statistics- Statistic
NFPA data shows ~174,000 highway vehicle fires per year in the US; EVs represent less than 1% of these despite growing market share- Excerpt
“"Data consistently demonstrates that despite media attention on EV fires, electric vehicles are substantially less prone to catching fire than traditional internal combustion engine vehicles." ”
- Source data from
- 2024-06-01
- Accessed
- 2026-04-18 · archived copy
- Calculation
- NFPA tracks all highway vehicle fires in the US and reports ~174,000 per year. EVs represent a growing but still small fraction of the fleet (~4% of new sales in 2023, ~2% of registered vehicles), and their share of total vehicle fires is well below their share of the fleet. This aggregate data confirms the direction of the per-100k comparison: EVs are underrepresented in vehicle fire statistics relative to their fleet share. The entry is included as a third data point to triangulate the US and Swedish per-vehicle comparisons.
- Independence
- NFPA fire incident reporting is an independent data pipeline from NHTSA/NTSB recall-based tracking and from Swedish MSB data.
-
[4] National Transportation Safety Board (NTSB) — Safety Risks to Emergency Responders from Lithium-Ion Battery Fires in Electric Vehicles (NTSB/SR-20/01)
Safety Risks to Emergency Responders from Lithium-Ion Battery Fires in Electric Vehicles (NTSB/SR-20/01)- Statistic
In all 3 of the crash-damaged EV batteries the NTSB studied, the battery reignited after firefighters extinguished the vehicle fire; one reignited at least six times and again 5 days after the crash, driven by stranded energy.- Excerpt
“"All three crash-damaged batteries reignited after firefighters extinguished the vehicle fires." "That stranded energy can cause a high-voltage battery to reignite multiple times after firefighters extinguish an electric vehicle fire. Emergency responders have no way of measuring how much energy remains in a damaged battery, and no way of draining that energy, other than such time-consuming methods as allowing a battery fire to burn itself out." ”
- Source data from
- 2020-11-13
- Accessed
- 2026-06-14 · archived copy
- Calculation
- NTSB Safety Report NTSB/SR-20/01 (adopted November 13, 2020) investigated all US domestic high-voltage EV battery fires the NTSB became aware of in a one-year window (August 2017–August 2018): three high-speed crash fires plus one noncrash fire. All three crash-damaged batteries reignited after the vehicle fire was extinguished — one reignited at least six times, including 5 days after the crash. This is the authoritative documentation of the reignition / stranded-energy phenomenon. It is a qualitative, mechanism-level finding from a four-case responder-safety investigation, NOT a population fire rate; it does not change the entry's headline ignition odds and is used only to substantiate the reignition caveat (per-incident severity, conditional on a fire already occurring).
-
[5] EV FireSafe (Australian fire-service research project) — EV Fire Reignition
EV Fire Reignition- Statistic
13% of EVs in EV FireSafe's reignition dataset reignited after initial suppression; one had 4 separate reignition events and another reignited 68 days after the incident; 5 reignited while on a tow truck.- Excerpt
“"Our research found 13% of the vehicles we studied reignited following initial suppression" "one vehicle had 4 seperate reignition events & another reignited 68 days after the initial incident" ”
- Source data from
- 2024-01-01
- Accessed
- 2026-06-14 · archived copy
- Calculation
- EV FireSafe is an Australian fire-service-funded research/advocacy project that compiles global EV fire incidents reported to it. Its 13% figure is the share of vehicles in its own reignition-incident dataset that reignited after suppression — a conditional rate (given a fire occurred) drawn from a self-selected fire-service sample with no published denominator, NOT a population probability and NOT comparable to the entry's per-100k ignition rate. It is cited only to put a number on the reignition phenomenon the NTSB documents qualitatively; treat the magnitude as indicative, not precise.
- Independence
- EV FireSafe's reignition dataset is collected independently of the NTSB's four-case US investigation, so the two sources cross-corroborate the reignition phenomenon from different fleets and methodologies — while sharing the limitation that neither is a population-representative reignition rate.