The aim of the study was to investigate the association between the use of benzodiazepine or benzodiazepine-like hypnotics and the risk of road traffic accidents. Data from three French national databases were matched: the health-care insurance database, police reports, and the police database of injury-related traffic accidents. A total of 72,685 drivers involved in injury-related road traffic accidents in France, from 2005 to 2008, were included in the study. The risk of being responsible for a traffic accident was higher in users of benzodiazepine hypnotics (odds ratio (OR) = 1.39 (1.08-1.79)) and in the 155 drivers to whom a dosage of more than one pill of zolpidem a day had been dispensed during the 5 months before the collision (OR = 2.46 (1.70-3.56)). No association was found between the use of zopiclone and risk of traffic accidents. Although this study did not find any association between the use of zolpidem as recommended and causation of traffic accidents, the potential risk related to possible abuse of the drug and risky driving behaviors should be further investigated. The results related to benzodiazepine hypnotics are consistent with those of previous studies.