Thirty-seven patients with previously untreated absence seizures were treated with ethosuximide. Seizures were completely controlled in 7 patients (19 percent); 90 to 100 percent control was achieved in 18 patients (49 percent) and 50 to 100 percent control in 35 (95 percent). Plasma ethosuximide concentration increased with dose, but variability in the plasma concentration produced by a given ethosuximide dose made it impossible to predict a patient's plasma concentration from the dose. The therapeutic range of plasma ethosuximide concentration was 40 to 100 mug per milliliter. Patients with evidence of structural central nervous system abnormalities responded as well or better to the drug as patients without such evidence. Ethosuximide did not impair psychometric performance, but rather resulted in improved performance in 17 cases. The side effects of ethosuximide were minor, and rarely required withdrawal of the drug.