One hundred psychiatric patients were carefully matched with 100 physically ill patients and their driving records compared. The psychiatric patients were consuming far greater quantities of psychotropic drugs and included a larger number of alcoholics and heavy drinkers. During the six months before admission there were no significant differences between the two groups of patients with respect to accident and traffic code infringements. Apart from individual patients, drugs did not appear to be influencing the outcome in statistical terms. Alcoholics and heavy drinkers showed an increased lifetime accident liability. No specific psychiatric diagnosis was otherwise associated with increased accident rates. The majority of accidents reported were relatively trivial.