An epidemiological community-based study of incident cases with non-provoked epileptic seizures, using case-referent methodology, was carried out to explore possible risk factors for epileptic seizures. 83 cases, between 17 and 74 years of age, of whom 67.4% had seizures of localized onset, were compared with 2 age- and sex-matched referents. Higher birth weight, movement disabilities, mental retardation, head trauma, brain tumor, depression, a period of unemployment during the previous 6 months and a history of epilepsy in relatives were more common in cases than in referent subjects. No difference was found in the socioeconomic factors investigated, except that the cases belonged to smaller households. Prematurity, home or hospital birth, parents' age at birth of cases or referents, febrile convulsions in relatives, various infections including meningitis and encephalitis, cerebrovascular disease, and alcohol, tobacco, sleep and nutritional habits were not found to be associated with development of seizures. The recent life events investigated, at home or at work, occurred as often in cases as in referents, except that significantly fewer cases had received any increase in salary during the last 6 months. The relationship between depression and development of seizures should be explored further. Moreover, the possibility of false negative results should be considered because of the sample size.