Among 600 drivers killed in single-vehicle crashes in North Carolina, alcohol was detected in 79.3% at usually high concentrations, while THC was found in only 7.8% and most drug-positive drivers also had high alcohol levels.
Read this if you want to understand the relative prevalence of alcohol versus cannabis in fatal crash data from the early 1980s.
Alcohol detected in 79.3% of 600 fatal single-vehicle crashes
What the researchers found
Researchers tested blood samples from all 600 drivers killed in single-vehicle crashes in North Carolina between 1978 and 1981 for alcohol, THC, barbiturates, cocaine, opiates, phencyclidine, amphetamines, and methaqualone.
Alcohol dominated the findings: detected in 79.3% of drivers, with 85.5% of alcohol-positive drivers having blood concentrations at or above 1.0 g/L (substantially above legal limits). Of all drivers tested, 67.8% had blood alcohol at or above this level.
THC was detected in 7.8% of drivers, methaqualone in 6.2%, and barbiturates in 3.0%. Other drugs were rarely detected. Critically, when drugs other than alcohol were found, the concentrations were usually within or below therapeutic ranges, and most drug-positive drivers also had high blood alcohol levels.
The researchers concluded that among the substances tested, alcohol was the only drug that appeared to have a significantly adverse effect on driving safety in this population.
Why it matters
This was one of the first large-scale studies to systematically test for cannabis alongside alcohol and other drugs in fatal crashes. Its finding that THC detection was relatively rare and usually accompanied by high alcohol levels helped contextualize cannabis's role in driving fatalities relative to alcohol.
The numbers in context
600 fatal crashes analyzed. Alcohol detected in 79.3% of drivers. THC detected in 7.8%. Methaqualone in 6.2%. Barbiturates in 3.0%. 67.8% of all drivers had blood alcohol at or above 1.0 g/L.
How the study worked
Comprehensive toxicology screening of blood specimens from an inclusive population of 600 single-vehicle operator fatalities in North Carolina, 1978-1981. All specimens were tested for ethanol, THC, barbiturates, cocaine, opiates, and phencyclidine. Subsets were tested for amphetamines and methaqualone.
What this study cannot tell us
Detection of THC in blood does not prove impairment at the time of the crash. THC metabolites clear the blood relatively quickly, so some cannabis-impaired drivers may have tested negative. The study could not isolate cannabis-only crashes because most THC-positive drivers also had alcohol. Blood collection timing after death may affect detection rates.
How to read the evidence
A comprehensive toxicology study of an inclusive population of fatal crashes. Strong methodology for prevalence data, but cannot establish causation.
When this study was published
Published in 1984, covering 1978-1981 data. Cannabis use rates, product potency, and drug testing sensitivity have all changed substantially.
The bigger picture
This study established an important baseline: alcohol was overwhelmingly the dominant intoxicant in fatal crashes. The relatively low THC detection rate and the confounding presence of alcohol in most drug-positive cases became recurring themes in decades of subsequent driving research.
Questions still open
- How many drivers used cannabis but tested negative due to rapid THC clearance? What would the results look like in a modern population with higher cannabis use rates? Can cannabis-impaired driving be identified independently of alcohol in crash data?
Common questions
Did cannabis cause any of these crashes?
Was alcohol the main factor?
Read the original research
Ethanol, marijuana, and other drug use in 600 drivers killed in single-vehicle crashes in North Carolina, 1978-1981.
Journal of forensic sciences, 29(4), 987-1026
Citation
Mason, A P; McBay, A J. (1984). Ethanol, marijuana, and other drug use in 600 drivers killed in single-vehicle crashes in North Carolina, 1978-1981.. Journal of forensic sciences, 29(4), 987-1026.