PURPOSE: We aimed to ascertain whether lumbar punctures should be done to prove febrile seizure, and to study what conditions can delay or abolish lumbar puncture. METHODS: This retrospective study was done for four years, from January, 1996 to December, 1999. The subjects were the patients who visited Masan Samsung Hospital for the first convulsion with fever, and whose age ranged from 1-month to 5-years old. The total number of patients was 197. We investigated the rate of meningitis, which was classified by age, season, convulsion type, associated symptoms, and results of blood test. All data were tested chi-square. RESULTS: The diagnosis rate of meningitis among all 197 patients was 7.6%, and of these only one patient was proved to have bacterial meningitis. The rate was high in the young(below 6 month), in males and in summer and autumn. Although the general appearance was good after resolution of the fever, the rate of meningitis was 6.5%. Patients with nausea and vomiting totalled 11%; 9.1% of them were diagnosed with meningitis. The diagnosis rate of meningitis in cases in which additional seizures took place were 33.3%. CONCLUSION: The ultimate goal of lumbar punctures in seizure patients with accompanying fever is to rule out bacterial meningitis, which, has been demonstrated to be rare. So we think that lumbar puncture can be postponed by close observation of pediatrics specialists and skilled nurses.