A multiple-dose administration study of exemestane (0.5-50 mg/day after breakfast for 7 days) was conducted in 32 normal healthy postmenopausal Japanese women using a single-blind, 3-step dose-titration method in order to investigate the safety, effect on serum concentration, amount of urinary estrogen excretion, and pharmacokinetics of the drug. Subjective/objective symptoms, in which a causal relationship with exemestane administration could not be excluded, were as follows: headache (8 cases: 1 each in the 0.5 and 25 mg groups, 2 in the 10 mg group and 4 in the 50 mg group), dizziness (2 cases: 1 each in the 0.5 and 25 mg groups), and fever (1 in the 25 mg group), all of which were mild and disappeared without treatment. The only abnormal laboratory findings were a mild increase in levels of GOT and GPT in 1 case and in the number of basophils in another case. There were no notable abnormal findings in vital signs, body weight or EKG. A dose-dependent decrease in serum estrogen level was observed between doses of 0.5 mg and 25 mg. The decrease was maximal at 25 mg, at which serum estrogen concentrations decreased to 14-27% of those observed at day 0. This decrease was maintained for one week, returning to baseline levels 2 weeks after the completion of drug administration. A similar result was also observed in the suppression of 24-hour urinary estrogen excretion. Exemestane was absorbed immediately after initial administration, reaching Cmax 0.9-2.6 hours post-administration. This was followed by a rapid decrease over the next 4-8 hours followed by a gradual decrease Cmax reached normal steady state values on day 5. Cmax and AUC0-24 values taken between administration of the first and final doses increased proportionally in a dose-dependent manner, suggesting that exemestane has a linear pharmacokinetic profile. Furthermore, results of the comparison of the trough concentrations of the initial dose with those of the final dose suggested no accumulative effects of the study drug.