We performed a randomized double-blind crossover therapeutic bioequivalency study of a generic (Epitol) versus a brand name (Tegretol) carbamazepine product under steady-state conditions in 40 epileptic patients. Each patient received 90-day supplies of Epitol or Tegretol and placebo, which replaced the usual dosage of the alternate product. Group A consisted of 20 seizure-free (from 5 months to 2 years) patients and group B of 20 patients with seizures refractory to drug therapy. In group A, four patients had seizures, two on both Epitol and Tegretol and two on Tegretol. In group B, the average seizure frequencies were 0.25 seizures per day on Epitol and 0.22 seizures per day on Tegretol. Average seizure frequencies were statistically the same (at a 20% difference, p less than 0.05). Areas under the curve were statistically the same (at a 20% difference, p = 0.05). Average peak heights were statistically the same (at a 20% difference, p less than 0.05). Average time to peak was earlier with Epitol. Epitol and Tegretol performed equally well in clinical efficacy and bioequivalency.