The efficacy of short-term diazepam prophylaxis in febrile convulsions was evaluated in a prospective, controlled study. A total of 289 consecutive children admitted with their first febrile seizure were randomized into two groups. One group received short-term prophylaxis for 18 months with rectally administered diazepam in solution whenever the temperature was greater than or equal to 38.5 degrees C. The control group received no prophylaxis, but diazepam rectally in the event of new seizures. The short-term prophylaxis, a mean of five doses of diazepam per child per year, afforded effective seizure control; the 18-month recurrence rate was reduced from 39% to 12% (P less than 0.001), the total number of recurrences from 77 to 23 (P less than 0.001), the long-lasting recurrences from 5.0% to 0.7% (P less than 0.05). The risk of subsequent epilepsy within the first 2 years was the same, regardless of receiving prophylaxis (3%) or not (3%); it was low after simple febrile convulsions (no cases of epilepsy in 230 children) but considerable after complex febrile seizures (20%) or seizures associated with severe interictal EEG abnormalities (50%).