In exploring the implications of hormonal influences on sexual behavior for reproduction, we have focused on androgens because of the convincing evidence for androgenic effects on female sexual motivation. We have been guided by the simple idea, based on clinical findings among hyperandrogenic women, that higher testosterone levels will increase female sexual motivation which in turn will result in increased coital frequency. However, careful consideration of the evidence fails to confirm such a role for testosterone in sexual behavior among normal women at any point in the reproductive span. While some amount of testosterone appears to be important in maintaining female sexual motivation, there is little evidence that variation in testosterone within the normal range is associated with variation in sexual motivation. Reported associations between testosterone and sexual behavior among married women cannot be interpreted as resulting from androgenic effects on sexual motivation. Thus we are left with the task of explaining why testosterone does not appear to play the same role in libido among hyperandrogenic and normal women. As mentioned earlier, some of the difference may attributable to the much higher levels of testosterone among hyperandrogenic women. Sherwin points out that the levels of sexual motivation decline with declining testosterone levels even while testosterone is well above normally occurring levels. It is possible that the brain is simply not sensitive to the variation in testosterone levels found in normal women. Other evidence suggests that the presence of intact ovaries may be equally important. A recent study of androgen replacement in naturally post-menopausal women failed to find a dramatic effect of testosterone on sexual motivation and behavior, despite levels similar to those in studies on surgically menopausal women, pointing to the importance of other factors associated with the presence of ovaries. A similar point can be made with regard to hyperandrogenism related to endogenous sources of testosterone, as in the case of PCOS. High levels of testosterone effectively disrupt ovarian function and interfere with other ovarian processes. Among hyperandrogenic women sexual behavior appears to be related to the direct effects of androgens on motivation, while the indirect effects of estrogen and progesterone are essentially eliminated. Among normal women, on the other hand, there is little evidence for such a dominating role of a direct androgenic effect on sexual behavior. Instead, other ovarian hormones, including estrogen and progesterone may also play a demonstrable role, despite the lack of strong evidence at this point.(ABSTRACT TRUNCATED AT 400 WORDS)