In order to study the role of drugs in driving, a responsibility analysis was developed to allow an assessment to be made of the driver's culpability or responsibility in an accident. Factors possibly mitigating drivers' responsibility in each accident were identified and scored. Factors considered were: condition of road, condition of vehicle, driving conditions, accident type, witness observations, road law obedience, difficulty of task, and level of fatigue. If a sufficient number of mitigating factors were identified a driver would be found to be either partly or totally exonerated from blameworthiness and scored either as a contributory or nonculpable driver. If drugs present in a driver contributed to accident causation, it would be expected that they would be overrepresented in culpable drivers, i.e. those drivers not exonerated from blame. A total of 341 driver fatalities occurring in Victoria were analysed for blood alcohol content (BAC). Twenty-nine percent had a BAC over .05% (the legal limit in Victoria). Alcohol-positive drivers were statistically overrepresented in the culpable group (p < .001), in single-vehicle accidents (p < .05) and those accidents in which vehicles left the road for no apparent reason (p < .001). Odds-ratio estimation of relative risk of culpable and nonculpable drivers showed that the relative risk rose disproportionately to BAC.