The outcome in 165 subjects with either an unknown (n = 93) or an alcohol-related (n = 72) seizure etiology, admitted to the emergency room of a general hospital in 1977-1978, was assessed after 10 years on the basis of subsequent hospital records and death-certificate-based mortality data. Alcohol and/or drug poisoning was the most frequent cause of death in the group with alcohol-related seizures. Sixty-four percent of the deaths in this group were directly related to alcohol abuse. The crude mortality was 45.8 (expected 8.6)/100 persons/10 years in the group with alcohol-related seizures and 15.1 (expected 6.0)/100 persons/10 years in the other group, the odds ratio between the groups being 4.8. Twenty percent of those with an unknown seizure etiology were found to show alcohol-related seizures, while the seizure etiology remained unknown in 59%, and a specific etiology other than alcohol abuse was revealed in 21% during the follow-up period. We conclude that alcohol abuse is an important, though often undetected, seizure etiology carrying a poor prognosis. The difference in mortality between the groups was due more to alcoholism than to seizures. There was no difference in mortality between those with a first alcohol-related seizure and those with previous alcohol-related seizures.