The issue of routine anticonvulsant prophylaxis for early and late posttraumatic epilepsy (PTE) has received much attention in the medical literature. Such problems as lack of standard definitions for early and late PTE, the retrospective design of most studies, the wide variability of inclusion and exclusion criteria, and the varied duration of follow-up make this body of literature extremely difficult to evaluate. Severe head trauma appears to cause injured neurons to become hyperexcitable; this in turn brings about the formation of an epileptogenic focus during the time between trauma and seizure occurrence. Both military and civilian head injury populations have been used to evaluate the incidence of PTE. Early seizures (i.e., less than 7 days) occur in approximately 3-5 percent of the head injury patients in both the military and civilian groups. Factors increasing this incidence include intracranial hematoma, focal neurologic deficits, posttraumatic amnesia (PTA) lasting greater than 24 hours, depressed skull fracture, and age less than 5 years. The incidence of late seizures is directly related to the extent of brain damage. The military population, composed primarily of cases with penetrating head injury, is associated with a late PTE incidence of approximately 30-50 percent. Closed head injuries in the military population involve a 5-15 percent seizure incidence. Late PTE incidence after head injuries in the civilian population is less than 5 percent. Risk factors associated with late PTE include loss of consciousness or PTA lasting greater than 24 hours, dural lacerations, depressed skull fractures, and various computerized tomography deficits. These factors vary slightly between the military and civilian populations.(ABSTRACT TRUNCATED AT 250 WORDS)