As part of a major undertaking to establish the contribution of drugs in road crashes in Quebec, the present study focuses on the role of cocaine. Coroner, forensic laboratory and police accident records from April 1999 to December 2000 were matched for 265 fatally injured drivers of passenger vehicles. Cocaine was found in 7.9% of urine samples and 6.0% of blood samples. In order to set up a control group, two roadside surveys were conducted in August 1999 and 2000. The survey sample was distributed proportionately to the number of fatal accidents per time of day and day of the week. During both daytime and nighttime, a total of 11,952 drivers participated in the two surveys among which 11,574 provided a breath sample (96.8%), 8,177 a saliva sample (68.4%) and 5,931 a urine sample (49.6%). Cocaine was detected in 1.1% of urine samples and 1.0% of saliva samples of the driving population. In both fatally injured drivers and driving population, cocaine was found mostly (> 90%) in four main types of combination:​ cocaine alone, cocaine + cannabis, cocaine + alcohol, cocaine + cannabis + alcohol. The data collected allowed two different analyses:​ a casecontrol (urine/urine) and a responsibility analysis (case-case approach) that compares cocaine cases to drug-free cases. Despite some data limitations, all analyses for the four main types of combination clearly suggest that cocaine use plays a role in fatal crashes.