Ethosuximide kinetics were determined in six normal healthy adults after a single dose (phase 1) and at steady-state (phase 2). After the completion of phase 2, valproic acid was added to the ethosuximide regimen (phase 3) to assess the possibility of drug interaction. Between phases 1 and 2 total clearance fell from 13.1 to 11.1 ml/hr/kg (P less than 0.05) and nonrenal clearance fell from 10.1 to 8.3 ml/hr/kg (P less than 0.05). When valproic acid was added (phase 3) there was no further change in total or nonrenal clearance (11.2 and 8.3 ml/hr/kg). To assess the possibility of nonlinear ethosuximide kinetics a review was conducted of patients who received ethosuximide as sole therapy for absence seizures. Of 106 patients, 10 met the required criterion that defined steady state. Data from seven of the 10 patients showed evidence of a nonlinear relationship when steady-state ethosuximide concentrations were plotted against dose.