What are the odds of causing an injury crash as an at-fault driver?
Evidence quality 4.13/5
Eight-dimension review score against the quality rubric . Each dimension scored 1–5.
- D1 Source grounding
- 4/5
- D2 Source authority
- 5/5
- D3 Arithmetic
- 3/5
- D4 Uncertainty
- 4/5
- D5 Scope
- 5/5
- D6 Prose
- 4/5
- D7 Perception honesty
- 4/5
- D8 Caveat completeness
- 4/5
Lifetime probability · lifetime, US adult
1 in 2.5
40% lifetime chance
Most people underestimate this.
range 1 in 3.6 to 1 in 1.8
● your factors — click this risk ▾ to reveal
≈ As likely as
Perceived
Drivers systematically underestimate their personal crash risk — a well-documented optimism bias in traffic safety research. Most licensed drivers believe they are above-average in skill, and the mental model of an "at-fault injury crash" tends to conjure images of reckless or drunk drivers, not the ordinary distracted or rushed driver that characterizes most real-world events. The actual lifetime probability of being the at-fault driver in a crash that injures someone is substantially higher than most people would guess.
Rough estimate: Most drivers would guess their lifetime odds of causing an injury crash are well below 1 in 10
Source: editorial intuition, not polled
Actual
~1 in 116 per year per US driver (injury crash involvement rate)
US licensed drivers (NHTSA CRSS 2023; all drivers involved in police-reported injury crashes)
Show derivation
NHTSA 2023 data: approximately 2.44 million people were injured in motor vehicle traffic crashes across an estimated 6.14 million total police-reported crashes. With 237.7 million licensed drivers (FHWA 2023), the per-year probability of a licensed driver being involved in any injury crash is approximately 2.44M / 237.7M ≈ 1.03% per year, or roughly 1 in 97. However, not all involved drivers are at fault; US crash statistics roughly split fault between the parties involved. Using a conservative 50% at-fault assumption yields an at-fault injury crash rate of approximately 1/116 per driver per year (adjusting for multi-vehicle crashes and single-vehicle crashes where the driver is always at fault). Compounding over 59 adult driving years: 1 − (1 − 1/116)^59 ≈ 0.40. The uncertainty range reflects that exact "at-fault" attribution is not uniformly captured in NHTSA police-reported data; the true at-fault involvement rate could plausibly range from 1 in 80 to 1 in 150 per year depending on attribution method.
Caveats: The 1/116 annual rate is derived from total injury-crash involvement divided by …
The 1/116 annual rate is derived from total injury-crash involvement divided by total licensed drivers, with a 50% at-fault adjustment applied. This is a population-level approximation; the actual proportion of crashes with a clearly designated at-fault driver varies by crash type, state law, and reporting practice. Some fraction of "injury crashes" involve only the at-fault driver as the injured party (single-vehicle run-off-road, etc.), where the financial and legal consequences differ from injuring a third party. The lifetime figure compounds annual independent exposures, which is a reasonable first-order model but does not capture the clustering of risk in early driving years or the dose-response of total mileage. The psychological consequence of injuring another person — separate from insurance, legal liability, or financial cost — is a real component of the risk that is not captured in any administrative statistic.
Risks at similar odds
Other risks with roughly the same likelihood — useful for calibration.
E-bike no helmet
What are the odds of serious head injury riding an e-bike without a helmet?
E-scooter no helmet
What are the odds of serious head injury riding an e-scooter without a helmet?
Trip disruption: war or disaster
What are the odds of a trip being significantly disrupted by war, political unrest, or natural disaster?
Driver kills pedestrian
What are the odds that a driver will kill a pedestrian in their lifetime?
Recently viewed on this device
Stored locally — clear anytime.
Pick challenger
Being the driver who causes a crash that injures someone else is one of the more statistically plausible adverse life events that most US drivers never think about. NHTSA’s 2023 data recorded approximately 2.44 million people injured in motor vehicle traffic crashes across 6.14 million total police-reported crashes — and with 237.7 million licensed drivers on US roads, the per-year involvement rate for an individual driver works out to roughly 1 in 116 for an injury crash where they bear some or all of the fault. Compounded over a 59-year driving lifetime, that translates to about a 1 in 2.5 chance of being the at-fault driver in at least one injury crash.
The consequences of an at-fault injury crash extend well beyond the immediate collision. Liability insurance covers third-party medical costs up to policy limits, but suits exceeding those limits, underinsured scenarios, license points, premium increases, and the psychological weight of having injured someone else are all real exposures. Drivers under 25 face roughly twice the national rate — a well-documented effect of inexperience compounded by higher rates of distracted and impaired driving in that age group. The optimism bias documented in traffic safety research runs strong here: most drivers rate themselves as above-average in skill, and the mental model of “the driver who causes a crash” tends to invoke reckless or impaired actors, not the ordinary momentary lapse that causes the median injury collision.
The lifetime figure here is a population average and hides enormous variation. A 60-year-old retired rural driver who puts on 4,000 miles per year faces a fundamentally different exposure profile than a 24-year-old urban delivery driver. National Safety Council data suggests medically consulted motor vehicle injuries may run as high as 4.9 million per year when counting crashes not reported to police — implying the police-report-based NHTSA figure, already substantial, likely understates the total event count. The risk is distributed across a lifetime of routine decisions: every commute, every late-night drive, every moment of divided attention.
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] National Highway Traffic Safety Administration (NHTSA) — Summary of Motor Vehicle Traffic Crashes: 2023 Data
Summary of Motor Vehicle Traffic Crashes: 2023 Data- Statistic
2.44 million people injured in motor vehicle traffic crashes in 2023; 6.14 million total police-reported crashes; up from 2.38 million injured in 2022- Excerpt
“"In 2023 an estimated 2.44 million people were injured in motor vehicle traffic crashes, compared to 2.38 million in 2022, an increase of 2.5 percent. The estimated number of police-reported traffic crashes increased from 5.93 million in 2022 to 6.14 million in 2023, a 3.5-percent increase." ”
- Source data from
- 2024-01-01
- Accessed
- 2026-05-14 · archived copy
- Calculation
- 2,440,000 injured persons / 237,700,000 licensed drivers (FHWA 2023) = 1.027% per driver per year of being involved in any crash that injured someone. Assuming roughly half of involved drivers are the at-fault party (a conservative approximation given that many single-vehicle crashes and many 2-vehicle crashes clearly assign fault to one driver), at-fault injury crash rate ≈ 0.86% per year. Using 1/116 ≈ 0.862%: 1 − (1 − 0.00862)^59 ≈ 0.40 lifetime probability.
- Independence
- NHTSA's CRSS (Crash Report Sampling System) is a probability sample of police-reported crashes; it is independent from insurance claims databases and court records. CRSS replaced NASS-GES as the national non-fatal crash data system beginning with 2016 crash year data.
-
[2] Bureau of Transportation Statistics / Federal Highway Administration — Licensed Drivers — Bureau of Transportation Statistics
Licensed Drivers — Bureau of Transportation Statistics- Statistic
237.7 million licensed drivers in the United States in 2023- Excerpt
“"According to the Federal Highway Administration, there were approximately 237.7 million licensed drivers in the United States in 2023, derived from state motor vehicle administration data submitted to FHWA's Highway Statistics Series." ”
- Source data from
- 2024-01-01
- Accessed
- 2026-05-14 · archived copy
- Calculation
- The licensed-driver denominator is used to convert total injury-crash counts (NHTSA) to a per-driver annual rate. FHWA's Highway Statistics DL-22 series provides the most comprehensive and consistently updated count of US licensed drivers by state, sex, and age group.
- Independence
- FHWA licensed-driver counts are based on state DMV administrative records and are wholly independent from NHTSA crash investigation data. The two datasets are combined here to derive a per-driver rate not published as such by either agency alone.
-
[3] National Safety Council — Motor Vehicle — Overview — Injury Facts
Motor Vehicle — Overview — Injury Facts- Statistic
Medically consulted motor vehicle injuries totaled 4.9 million in 2024; 39,345 estimated traffic fatalities in 2024; odds of dying in a motor vehicle crash are 1 in 101 lifetime- Excerpt
“"Medically consulted injuries in motor-vehicle incidents totaled 4.9 million in 2024. The National Safety Council estimates the odds of dying in a motor vehicle crash at approximately 1 in 101 over a lifetime for the average American." ”
- Source data from
- 2025-01-01
- Accessed
- 2026-05-14 · archived copy
- Calculation
- The NSC medically consulted injury figure of 4.9 million in 2024 is higher than NHTSA's 2.44 million police-reported injured count because it includes crashes not reported to police and injuries with delayed medical presentation. It is cited here for context on the upper bound of total injury events, not for the primary rate calculation, which uses the police-reported NHTSA figure.
- Independence
- NSC compiles injury data from insurance payors, employer records, and medical billing datasets, making it methodologically distinct from both NHTSA's police-report sampling system and FHWA's DMV administrative records.