A suppository for rectal administration of carbamazepine has been developed for situations in which it is unsuitable to use the oral route of administration. In an open, controlled, within-patient study, the pharmacokinetics, clinical efficacy, and tolerability of carbamazepine slow-release tablets were compared with those of carbamazepine suppositories in children with epilepsy. The pharmacokinetic part of the study comprised 22 children, and an additional nine children were included in the clinical part of the study. Treatment with slow-release tablets was replaced for 7 days with carbamazepine suppositories in bioequivalent dosage. Clinical factors such as the rate of seizures and the local tolerability were studied, and an overall assessment of efficacy was made. In the pharmacokinetic part, 24-hour plasma concentration curves for carbamazepine and carbamazepine-10,11-epoxide were recorded. The plasma concentration profiles (minimum, maximum, and mean concentrations, fluctuation index, and area under the curve) for carbamazepine and the other metabolites did not show any significant differences between oral and rectal administration when the suppository dose was increased by 25% compared to the tablets. No increase in seizure frequency was detected, and the overall assessment was very good to good in 25 of the 29 epileptic children. Increased flatulence during treatment with suppositories was noted in two children, one had anal irritation, and one had nausea/vomiting. Treatment with carbamazepine slow-release tablets in children with epilepsy can be replaced by carbamazepine suppositories in 25% higher dosage, with good clinical effect and appropriate pharmacokinetic values, when it is unsuitable to use the common oral route of administration.