Every year, the causes behind car crashes shift in ways that are easy to miss until you look at the evidence side by side. In the U.S., the fatality rate fell from 12.9 per 100,000 in 2011 to 10.6 by 2022, even as speeding, alcohol impairment, distraction, fatigue, and restraint failures continue to drive disproportionate harm. This post connects those threads with specific findings, from unbelted occupant deaths and alcohol levels in fatal pedestrian crashes to how distracted and sleep-deprived driving measurably changes crash risk.
Key Takeaways
- In a U.S. NHTSA analysis, unbelted passenger vehicle occupants account for 43% of passenger vehicle occupant deaths
- In low- and middle-income countries, 24% of road deaths are among pedestrians (reflecting exposure and crash causation contexts)
- The OECD estimates that road fatalities are strongly linked to driver behavior; in OECD countries, alcohol and speed remain major contributors according to road safety reviews (quantified in OECD road safety outlook)
- 26% of crash fatalities in the U.S. were linked to speeding-related factors
- In a large meta-analysis of human factors studies, the odds of being involved in a crash increased by 2–3x when drivers used hand-held devices
- A systematic review found that distracted driving (visual-manual tasks such as texting) was consistently associated with increased crash risk; one pooled estimate showed ~2x increased risk compared with undistracted driving
- A NHTSA report estimated that preventing alcohol-impaired driving could reduce fatalities by about 7,000–10,000 annually (quantified scenario)
- A systematic review reported that speed management interventions (e.g., speed cameras) can reduce speed and reduce injury crashes, with quantified effects across included studies
- A Cochrane review on speed cameras found reductions in injury collisions with quantified pooled effects
- In the U.S., 49% of speeding-related fatalities in 2022 involved vehicles traveling 10+ mph over the speed limit (FARS-based estimate)
- The WHO Global Status Report on Road Safety 2018 estimated 1.35 million road deaths in 2016 globally (baseline total)
- In the U.S., the fatality rate per 100,000 population was 12.9 in 2011 and fell to 10.6 by 2022 (trend showing overall outcomes amid cause-specific efforts)
- A U.S. case-control study found that texting while driving increased the odds of crash involvement (odds ratio reported in the study)
- A systematic review in Accident Analysis & Prevention reported that fatigue-related impairments increase near-crash/crash risk by roughly 1.3 to 2x depending on study design
- A meta-analysis in Traffic Injury Prevention reported that drivers who were distracted showed higher crash involvement risk than nondistracted drivers, with a pooled effect near ~1.5–2x
Seat belts, speed control, and reducing alcohol and distracted driving could prevent many road deaths annually.
Related reading
02 · Category
Causation Percentages6 stats
01
26% of crash fatalities in the U.S. were linked to speeding-related factors
02
In a large meta-analysis of human factors studies, the odds of being involved in a crash increased by 2–3x when drivers used hand-held devices
03
A systematic review found that distracted driving (visual-manual tasks such as texting) was consistently associated with increased crash risk; one pooled estimate showed ~2x increased risk compared with undistracted driving
04
A study of U.S. hospital trauma registry data found that 52.6% of pedestrian crash victims had alcohol intoxication markers
05
A peer-reviewed review reported that fatigue increases crash risk by about 2x (sleep-related factors) in driving populations
06
In a U.S. case-crossover study of crashes, blood alcohol concentration ≥0.08% was present in 40% of fatal pedestrian crashes
Interpretation
Causation Percentages Interpretation
Overall, these causation percentages show that multiple modifiable driver and road-user factors meaningfully raise crash risk, with intoxication showing up in 52.6% of injured pedestrians and at least 40% of fatal pedestrian crashes involving a BAC of 0.08% or higher.
03 · Category
Policy & Intervention Impacts7 stats
01
A NHTSA report estimated that preventing alcohol-impaired driving could reduce fatalities by about 7,000–10,000 annually (quantified scenario)
02
A systematic review reported that speed management interventions (e.g., speed cameras) can reduce speed and reduce injury crashes, with quantified effects across included studies
03
A Cochrane review on speed cameras found reductions in injury collisions with quantified pooled effects
04
A Cochrane review reported that graduated driver licensing reduces crash risk for novice drivers, with quantified effect estimates
05
A peer-reviewed study found that higher minimum drinking ages reduced alcohol-related traffic crashes by a quantified percent (historical policy impact)
06
A meta-analysis found that speed camera programs reduce injury crashes by a quantified percent compared with areas without cameras
07
A systematic review of alcohol ignition interlocks reported reductions in repeat offenses by about 40% to 60% depending on jurisdiction (quantified in review)
Interpretation
Policy & Intervention Impacts Interpretation
Overall, the policy and intervention evidence shows that targeted measures can make a measurable dent in crashes, including alcohol-impaired driving prevention cutting fatalities by about 7,000 to 10,000 per year and speed management like speed cameras reducing injury crashes by pooled effects across studies while graduated driver licensing lowers novice crash risk and ignition interlocks reduce repeat offenses by roughly 40% to 60% depending on jurisdiction.
More related reading
04 · Category
Trends & Totals8 stats
01
In the U.S., 49% of speeding-related fatalities in 2022 involved vehicles traveling 10+ mph over the speed limit (FARS-based estimate)
02
The WHO Global Status Report on Road Safety 2018 estimated 1.35 million road deaths in 2016 globally (baseline total)
03
In the U.S., the fatality rate per 100,000 population was 12.9 in 2011 and fell to 10.6 by 2022 (trend showing overall outcomes amid cause-specific efforts)
04
In the U.S., 8,210 people were killed in crashes involving speeding in 2019 (DOT/NHTSA estimate)
05
In Brazil, road traffic injuries caused about 31,000 deaths in 2019 (WHO Global Health Estimates), providing regional outcome totals for cause modeling
06
In the EU, 25,500 people died in 2021 on EU roads (European Commission summary), enabling trend analysis
07
In Australia, 1,178 people died on roads in 2022 (Australian government road safety data), establishing baseline totals by year
08
In Canada, 2,981 road deaths occurred in 2022 (Statistics Canada), baseline totals for cause analysis
Interpretation
Trends & Totals Interpretation
Across countries, road death tolls remain substantial while outcomes show measurable improvement in some places, such as the U.S. fatality rate dropping from 12.9 per 100,000 in 2011 to 10.6 by 2022, even though 8,210 people were still killed in speeding-related crashes in 2019.
05 · Category
Injury Risk Multipliers13 stats
01
A U.S. case-control study found that texting while driving increased the odds of crash involvement (odds ratio reported in the study)
02
A systematic review in Accident Analysis & Prevention reported that fatigue-related impairments increase near-crash/crash risk by roughly 1.3 to 2x depending on study design
03
A meta-analysis in Traffic Injury Prevention reported that drivers who were distracted showed higher crash involvement risk than nondistracted drivers, with a pooled effect near ~1.5–2x
04
A study published in Epidemiology (alcohol) reported substantially elevated odds of crash involvement at higher BAC levels compared with 0.0
05
NHTSA has reported that seat belts reduce serious injury by about 50% for certain front-seat occupant groups (restraint effectiveness)
06
A Cochrane review on helmet effectiveness reported that motorcycle helmets reduce the risk of death and head injury (quantified)
07
A study in JAMA Network Open reported that alcohol use among drivers is associated with a higher risk of severe crash outcomes (odds ratios quantified)
08
A study in the American Journal of Epidemiology reported that drivers with recent cannabis use had elevated crash risk (odds ratio reported)
09
A meta-analysis found that daytime running lights can increase visibility; however, crash causation impact varies and the pooled effect is quantified (collision/incident reduction)
10
In a vehicle crash study, making lane changes with turn-signal omission was associated with increased conflict risk (quantified odds/risk ratios)
11
A driving simulator study reported reaction-time decrements under distraction, with measured delays (e.g., hundreds of milliseconds) affecting crash avoidance performance
12
A study in Sleep Medicine Reviews reported that sleep restriction of ~24 hours without sleep impairs driving performance with measurable lane-keeping deterioration
13
A meta-analysis in Accident Analysis & Prevention estimated that reducing speeding compliance gaps can lower crash fatalities by a quantifiable percent depending on policy design
Interpretation
Injury Risk Multipliers Interpretation
Across injury risk multipliers, multiple evidence sources show that behaviors impairing attention or judgment raise crash involvement and severity by about 1.3 to 2 times, while effective restraints like seat belts and helmets can cut serious injury or death by roughly half or more.
Reference
Cite This Report
This report is designed to be cited. We maintain stable URLs and versioned verification dates. Copy the format appropriate for your publication below.
APA
Marcus Engström. (2026, February 13). Car Accident Causes Statistics. Gitnux. https://gitnux.org/car-accident-causes-statistics
MLA
Marcus Engström. "Car Accident Causes Statistics." Gitnux, 13 Feb 2026, https://gitnux.org/car-accident-causes-statistics.
Chicago
Marcus Engström. 2026. "Car Accident Causes Statistics." Gitnux. https://gitnux.org/car-accident-causes-statistics.
Sources & references
38 datasets cited across this report · attribution is report-level
+21 additional datasets cited (not shown individually)