We compared the therapeutic efficacy of carbamazepine versus sulthiame in patients with benign childhood epilepsy with centrotemporal spikes. Drug efficacy was evaluated only in those patients who initiated treatment with any drug after at least three seizures. Thirty-eight patients who received carbamazepine and 18 patients who received sulthiame were included in the analysis. Cessation of seizures was observed in 73.6% of the former and in 66.7% of the latter (P = not significant). Five of eight patients who were switched to sulthiame after failing carbamazepine became seizure free, whereas none of the three patients who failed sulthiame became seizure free after being switched to carbamazepine. The rate of drug discontinuation owing to adverse reaction was 15% in carbamazepine and 14.3% in sulthiame. Normalization of interictal epileptiform activity on electroencephalography was seen more often following treatment with sulthiame (71%) than with carbamazepine (42%) (P = not significant). No significant differences between these two medications were found in the treatment of benign childhood epilepsy with centrotemporal spikes in this small patient sample.