Purpose: Although there is strong circumstantial evidence that androgens are implicated in the etiology of prostate cancer, epidemiologic investigations have failed to demonstrate consistently that one or more steroid hormones are implicated. In contrast, recent epidemiologic studies unequivocally link serum insulin-like growth factor 1 (IGF-1) levels with risk for prostate cancer. Methods: We have performed the first meta-analysis of all previously published studies on hormonal predictors of risk for prostate cancer. Results: A meta-analysis restricted to studies that performed mutual adjustment for all measured serum hormones, age, and body mass index indicated that men whose total testosterone is in the highest quartile are 2.34 times more likely to develop prostate cancer (95% confidence interval, 1.30 to 4.20). In contrast, levels of dihydrotestosterone and estradiol do not seem to play a role of equal importance. The only study that provides multivariably adjusted sex hormone- binding globulin data indicates that this binding protein is inversely related to prostate cancer risk (odds ratio, 0.46; 95% confidence interval, 0.24 to 0.89). Finally, all three studies that examined the role of serum IGF-1 have consistently demonstrated a positive and significant association with prostate cancer risk that is similar in magnitude to that of testosterone. Conclusion: Men with either serum testosterone or IGF-1 levels in upper quartile of the population distribution have an approximately two- fold higher risk for developing prostate cancer. (C) 2000 American Society of Clinical Oncology.