Spermatogenesis, reproductive luminal contents and androgen concentrations were examined in hypophysectomized male rats treated with 1 of 3 testosterone (T) dosages for 60-64 days and in sham-operated controls. Hypophysectomized rats treated with 2 cm long T implants showed normal mating but reduced fertility, while normal fertility was maintained in animals given 8 or 3 .times. 8 cm T. Spermatogenesis in the hypophysectomized groups treated with 2 cm T for 10 days was generally arrested at the spermatocyte stage, while in the hypophysectomized animals treated with 2 cm T for 64 days spermatogenesis was halted at the spermatid stage. The 8 and 3 .times. 8 T-treated hypophysectomized animals had normal spermatogenesis with only minimal focal areas of degenerating seminiferous epithelium. Serum T concentrations were reduced by hypophysectomy, maintained at control levels by the 2 cm T dosage and increased to pharmacological levels in a dose-dependent manner by 8 and 3 .times. 8 cm T treatments. Testicular T concentrations also responded in a dose-dependent manner but the 3 .times. 8 cm T dose was not sufficient to keep testicular T at control levels. Dihydrotestosterone (DHT) in the caput epididymidis was maintained at control levels by the 8, reduced by the 2 and elevated by the 3 .times. 8 cm T treatment. Without gonadotropins, higher than normal levels of serum T are required to maintain normal fertility, although this normal reproductive capacity is possible even with greatly reduced testicular concentrations.