To determine the efficacy of pyridoxine in treating seizures, 90 infants and children with recurrent convulsions primarily due to acute infectious diseases were enrolled in the present study. Forty patients were treated with high-dose pyridoxine (30 or 50 mg/kg/day) by intravenous infusion, and 50 subjects served as controls. Antiepileptic drugs and other therapies were similar in the two groups except for pyridoxine. Clinical efficacy criteria were based on the frequency of convulsions per day and on the duration of individual seizures after therapy was initiated. The results indicated that total response rates in the pyridoxine group and control group were 92.5% and 64%, respectively (chi-square = 14.68, P < .001). After initiation of therapy, seizures resolved after 2.4 +/- 1.4 days in the pyridoxine group and after 3.7 +/- 2.0 days in the control group (t = 3.67, P < .001). No adverse effects of pyridoxine were apparent during the observation period. We conclude that pyridoxine is an effective, safe, well-tolerated, and relatively inexpensive adjunct to routine antiepileptic drugs for treatment of recurrent seizures in children.