A multicenter cooperative study was conducted to investigate factors influencing posttraumatic epilepsy (PTE) and to evaluate the prophylactic effect of anticonvulsants. A retrospective study of 102 PTE patients revealed the following typical clinical features: occurrence in young males, traffic accidents, contusion and/or cerebral hematoma. The latent period after the injury was longer in children. The percentage of EEG paroxysmal activity gradually increased as the generalized abnormality diminished. A retrospective-prospective study of 1998 patients who suffered a head injury between 1984 and 1988 was conducted till 1994. During the follow-up period, 62 patients (3.1%) developed PTE. The drop-out cases were excluded, and the 154 cases followed at least two years were analyzed. Statistical analysis of differences between patients with and without PTE suggested following factors: young, immediate early epilepsy (within 24 hours after injury; IMEE) and early epilepsy (within one week after injury). The risk with the highest relative risk rate was early epilepsy. Multiple regression analysis revealed that three factors, IMEE, early epilepsy and young age, contributed to the prediction of PTE. There was no significant difference in the percentage of patients having PTE in the group treated with anticonvulsants and the untreated group. Anticonvulsant treatment after head injury was unlikely to have a prophylactic effect on the development of PTE.