Recent studies suggest that the antiprogestin RU486 may have a direct effect on human ovarian luteal function. To further examine this possibility, we have studied the effect of RU486 on ovarian steroidogenesis using human granulosa cells obtained from women undergoing in vitro fertilization. RU486 at concentrations of 0.1, 2, 5, 10 and 100 nM was incubated with 10(5) granulosa cells over 72 hours. Significant suppression of progesterone production occurred following treatment of cultured cells with 2, 5, 10, and 100 nM RU486 at 24 hours (p less than 0.05) and 48 hours (p less than 0.01). At 72 hours, significant decreases in progesterone production were observed with 10 nM (p less than 0.05) and 100 nM RU486 (p less than 0.01). The greatest effect of RU486 on progesterone production occurred at 24 hours of incubation (slope = -8.03) compared with 48 (slope = -4.71) or 72 (slope = -2.31) hours (p less than 0.01). Maximal suppression of progesterone production occurred using 10 nM RU486 with no further significant suppression observed with 100 nM RU486. Other steroids (R5020, DHA) failed to suppress progesterone production suggesting that the observed inhibitory effect on progesterone was specific to RU486. To better understand how RU486 decreases progesterone production in granulosa cell cultures, we measured human ovarian 3B-hydroxysteroid dehydrogenase (3BHSD) in the presence and absence of RU486 in vitro. A significant dose-dependent decrease in the activity of 3BHSD was observed at concentrations of RU486 that were equal to or greater than substrate concentration. Taken together, these findings suggest that RU486 may directly affect human ovarian progesterone production through a mechanism that involves a reduction in 3BHSD activity.