Male rats were castrated and treated with various doses of testosterone (T) or 5 alpha-dihydrotestosterone (DHT) by implantation of steroid releasing polydimethylsiloxane depots. After 14 days, the animals were killed and the concentration of T, DHT and 5 alpha-androstane-3 alpha,17 beta-diol (Diol) was measured by radioimmunoassay in plasma, in the homogenates of the prostate (PR), the bulbocavernosus/levator ani muscle (BCLA), skeletal muscle (SM) and heart muscle (HM), as well as in the cytosol and the nuclear fraction of these tissues. Additionally, the weight and DNA content of PR and the levator ani muscle were measured. The main results were: After T or DHT administration the main androgens in plasma were T or Diol, respectively, the concentration of each being linearly dependent from the administered steroid dose (T after T-: r = 0.97, P less than 0.001. Diol after DHT administration: r = 0.97, P less than 0.001). T in a dose of 60 micrograms/24 h maintained the concentrations of T, DHT and Diol in plasma in the range of untreated controls. Irrespective of the T doses administered, its concentration in PR tissue was comparatively low. T administration resulted in a significant accumulation of DHT in PR, when compared to plasma T. The steepest increase of DHT in PR was found with low doses of T administered (0.6-12 micrograms/24 h). At higher doses (12-600 micrograms/24 h) the further increase in the tissue was rather parallel to that of plasma T. A comparable picture was found for the DHT concentration in cytosol and in nuclei. The Diol concentration in the cytosol was linearly related to the administered T dose (r = 0.80, P less than 0.001). This steroid could not be detected in nuclei (less than 0.5 pmol/mg DNA). Administration of DHT in principle mimicked the DHT data observed after T treatment. In muscle homogenates no accumulation of neither T, DHT nor Diol was observed. In the nuclear fractions only after the highest doses of T measurable amounts of T and DHT were found. Whereas a positive correlation between DHT values in nuclei and weight increases in prostate was present, no such parallelism could be demonstrated in muscular tissue. The data are discussed with respect to the present knowledge of androgen binding and metabolism in different organs and to a possible specific role of nuclear DHT accumulation in mediating androgen dependent processes.