In recent years, reproductive disorders in athletes have been observed in increasing numbers and proportions as larger numbers of women are participating in more strenuous athletic activities. Clinical problems with infertility and decreased bone density have been reported in this group, which has prided itself on its health. Hypoestrogenic women, including amenorrheic athletes, should definitely seek medical treatment for the prevention of osteoporosis, but before women sacrifice the many benefits of strenuous exercise, it is appropriate to consider the strength of the evidence that exercise training in particular, rather than other, extraneous aspects of athletic training, is responsible for the hormonal changes associated with reproductive disorders. This review has examined this first link in the postulated causal chain between exercise and these clinical conditions. In this review, only those studies that have compared trained or highly trained women with untrained or less-trained women, either cross-sectionally or longitudinally have been considered. In assessing the credibility of the reported results, we have found that the extensive precautions, painstaking techniques, and powerful protocols required for studying the question have only slowly been appreciated and never fully implemented by investigators. As a result, most of the published results are of little value. The few studies less susceptible to such criticism are consistent in their observations of gonadotropin and ovarian steroid suppression in women participating in athletic training. Nevertheless, even these results may have been biased by confounding factors, and thus at the present time the first link in the causal chain of events from exercise training to infertility and fractures cannot be regarded as proven. If the observed suppression of reproductive hormones is, indeed, due to exercise training, it may be dependent upon either the intensity of exercise training or the rate at which the intensity is increased. A dependence upon the rate at which exercise intensity is increased may explain why large proportions of athletes at all exercise intensity levels remain eumenorrheic. If future studies eliminate the biasing effects of known confounding factors and confirm that the female reproductive system can be suppressed by exercise training in particular, then additional research will be needed to determine the training intensity or rate of change in intensity required for these effects to occur. Research will also be needed to determine how gynecologic age, body composition, diet, and psychologic status may exaggerate or attenuate these effects.(ABSTRACT TRUNCATED AT 400 WORDS)