The clarification of the mechanisms of action of cannabis and its effects on motor and cognitive functions, the results of previous studies performed on driving simulators and closed or open-road driving trials, are important criteria for highlighting the increased risk of road crashes for drivers after a recent use of cannabis. In addition epidemiological studies allow to measure the magnitude of the problem. A survey of French epidemiological studies performed from 1999 to 2004, as well as the data of THC distribution in tissues studies performed on man and animal allowed us to draw a number of conclusions. The risk of road crash after a recent use of cannabis is increased by more than 2.4 in all studies. The prevalence of cannabis use in drivers involved in a road crash has dramatically increased during the last years. For methodological reasons (a too high threshold for THC positivity, a too long time delay between accident and blood sampling), the annual number of fatal cases induced by a cannabis use was likely underestimated. This assessment is consistent with recent data which indicate that THC could be still present in brain while absent in blood. A positivity threshold for THC in blood of 0.5 ng/mL would be more appropriated. So, all recent French studies highlighted that a recent use of cannabis impairs driving ability and that it would be advisable to intensify roadside testing for drugs of abuse.