The aim of this review is to identify circulating steroids associated with prostate cancer (PCa) in bibliographic reports. First, we summarize chronological reports comparing circulating steroid levels in men with a high risk of PCa, such as African-Americans (AAs), with men having a lower risk. Higher testosterone plasma levels in young AA men have been reported. However, the difference between AAs and Caucasians decreases with age. When measured, dihydrotestosterone (DHT) was found to be higher in AAs; however, these results must be taken with caution, as immunological assay of this steroid is difficult. Second, we summarize chronological reports concerning circulating steroids assayed in blood samples drawn several years before a diagnosis of PCa was made, compared to controls. These nested case-control studies did not lead to straightforward conclusions regarding an increase in circulating testosterone. However, large collaborative studies showed a trend of a decrease in plasma sex hormone binding globulin, consequently an increase in bioavailable plasma testosterone in PCa. These nested case-control studies failed to associate plasma DHT levels with PCa risk. Third, we summarize numerous chronological publications relating plasma concentrations, measured at the time of PCa diagnosis. Numerous results showed a decrease in plasma testosterone levels in aggressive PCa (high Gleason score, advanced stage with positive surgical margins). Testosterone levels returned to normal several months after prostatectomy. All plasma hormone assay levels were always carried out using immunological methods. We recommend replacing this methodology by mass spectrometry coupled with gas chromatography or liquid chromatography in these epidemiologic studies.