Although schedule-dependent cytotoxicity of etoposide has been reported, its mechanisms have not been elucidated fully. In this study, we attempted to clarify what concentration, time or exposure dose (ED, concentration of drug x time) of etoposide result in the antitumor effect on human ovarian cancer cells from the standpoint of cell cycle perturbation. The different ED were produced by different drug treatment schedules: 10 microgram/ml x 4 h (ED 40), 1.66 microgram/ml x 24 h (ED 40), 5 microgram/ml x 4 h (ED 20), 0.83 microgram/ml x 24 h (ED 20), 10 microgram/ml x 0.8 h (ED 8), 5 microgram/ml x 1.6 h (ED 8), 2 microgram/ml x 4 h (ED 8), 0.33 microgram/ml x 24 h (ED 8). Cell cycle perturbation on day 5 and the suppression of cell growth were dependent on the drug concentration at the lowest exposure dose (ED 8), but were dependent on ED at the higher EDs (20 or 40). The percentage of cells in the G2/M fraction significantly decreased from day 5 to day 7 in BG-1 cells treated at ED 20 or treated with higher concentrations (10 and 5 microgram/ml) at ED 8, but not in those treated at ED 40 or treated with lower concentrations (2 and 0.33 microgram/ml) at ED 8. Therefore, the cytotoxic mechanism of etoposide may not be explained by simple schedule dependency.