Case-crossover and case-series analyses are 2 epidemiologic approaches that can be used to evaluate the association of exposures with acute events. Using a primary care database from the United Kingdom and these 2 statistical approaches, the authors investigated the impact of using benzodiazepines, nonbenzodiazepine hypnotics, beta-blockers, selective serotonin reuptake inhibitors, tricyclic antidepressants, opioids, and antihistamines on the risk of motor vehicle crashes in 1986-2004. For 49,821 individuals aged 18-74 years, involvement in a motor vehicle crash was documented. The outcome of the case-crossover analyses varied according to the choice of control period, so the case-series approach was preferred. The first 4 weeks of treatment with a combined acetaminophen and opioid preparation was associated with an increased risk of motor vehicle crash (incidence rate ratio = 2.06, 99% confidence interval: 1.84, 2.32), as was use of an opioid alone (incidence rate ratio = 1.70, 99% confidence interval: 1.39, 2.08) and benzodiazepines (incidence rate ratio = 1.94, 99% confidence interval: 1.62, 2.32). Use of selective serotonin reuptake inhibitors, nonbenzodiazepine hypnotics, and antihistamines for more than 4 weeks was associated with motor vehicle crash, but shorter term use was not. The results obtained are broadly consistent with those from well-designed case-control studies and demonstrate how case-only techniques optimize the use of routinely collected data for epidemiologic studies.