The administration of exogenous testosterone (T) to eugonadal men causes suppression of gonadotropin secretion and thus of spermatogenesis. This is currently being investigated as a possible method of hormonal male contraceptive, but complete suppression of spermatogenesis to azoospermia is induced in only 50-70% of Caucasian men; the remainder maintain a low rate of spermatogenesis. The basis for this polymorphism in response is unclear. The enzyme 5 alpha-reductase (5 alpha R) converts T to dihydrotestosterone (DHT) and is important in determining the magnitude of the androgen stimulus in some tissues. We investigated whether the maintenance of spermatogenesis in men remaining oligozoospermic while receiving suppressive doses of T is associated with evidence of increased 5 alpha R activity. Thirty-three normal men were given 200 mg T enanthate (TE), im, weekly in a clinical trial of hormonal male contraception. The MCR of T (MCRT) and the conversion ratio of T to DHT (CRT-DHT) were measured by infusion of [3H]T, plasma levels of DHT and androstanediol glucuronide (AdiolG) were measured by RIA, and 24-h urinary steroid metabolites were measured by capillary column gas chromatography. Sperm density decreased in all men; 18 achieved azoospermia by 20 weeks of treatment, and the remainder had a mean sperm density of 2.0 +/- 0.8 x 10(5)/mL at that time. This treatment caused increases in plasma T levels and MCRT, but with no differences between azoospermic and oligozoospermic responders. There were no differences in CRT-DHT plasma DHT, or AdiolG before treatment, but after 16 weeks, CRT-DHT had increased in the oligozoospermic responders, but not in the azoospermic responders. TE treatment increased plasma DHT and AdiolG levels in both groups, but the increases in both 5 alpha R metabolites were significantly greater in the oligozoospermic responders. Urinary excretion of etiocholanolone and androsterone was increased after 16 weeks of TE treatment, but did not differ between the two groups, andetiocholanolone/androsterone ratios did not differ greatly from unity. There was no change in urinary excretion of tetrahydrocortisol, allo-tetrahydrocortisol, or cortisone after 16 weeks of TE treatment in either group. These results suggest that after TE administration there is a selective increase in 5 alpha R activity in those men who remain oligozoospermic, but not in those becoming azoospermic. This difference in the androgenic milieu may underlie the incomplete suppression in the oligozoospermic responders, in whom a low rate of spermatogenesis is maintained despite the apparent absence of gonadotropins.