The pharmacokinetics and relative bioavailability of fluoxetine capsules (reference) and tablets (test) were compared in 24 healthy subjects of both sexes after a single 20 mg oral dose of fluoxetine (as a hydrochloride salt). A randomized, crossover design with a 2-week wash-out period between each dose was applied. Serum samples, obtained before dosing and at various appropriate time points up to 192 hours, were analyzed for fluoxetine and norfluoxetine content by a simple, accurate and precise HPLC method. ANOVA, power analysis, 90% confidence intervals (CI), and two one-sided tests were used for the statistical analysis of pharmacokinetic parameters. The tolerability of the preparations was good. The respective point estimates of the ratios of the geometric means of log-Cmax and log-AUC(0-infinity) of fluoxetine were 0.912 and 0.935 with 90% of 0.838-0.992 and 0.857-1.020. The corresponding point estimates of norfluoxetine were 0.952 (90% CI = 0.843-1.075) and 0.904 (90% CI = 0.807-1.013), respectively. Since both 90% CI for the AUC(0-infinity). and Cmax geometric mean ratios of fluoxetine and norfluoxetine were included in the 80% to 125% interval proposed by the FDA the test drug (fluoxetine tablets) was considered bioequivalent to the reference one (Prozac capsules) according both to the rate and extent of absorption.