The literature has often suggested a relationship between psychiatric illness and increased risk for motor vehicle accidents but few data exist, particularly those from prospective or case-controlled studies. The present study examined detailed accident and injury data from a large series (N = 1778) of motor vehicle crashes that included persons with diagnoses of psychiatric illnesses (n = 17); matched controls were also studied. As expected, drivers with psychiatric diagnoses used psychotropic medications more frequently than did controls. Drivers with psychiatric diagnoses did not have more frequent single-car crashes, unsafe speed or failure to yield violations, or less restraint use than did controls. Such drivers did less often drive motorcycles and tended to be less often cited for alcohol use. Implications of these results for efforts toward prevention and for further study are discussed.