This study aimed to establish whether the degree of suppression of serum FSH and LH was related to sperm concentration in three testosterone (T) plus progestin contraceptive regimens. We measured serum FSH and LH using a modified, highly sensitive immunofluorometric assay in samples obtained from three published studies using T enanthate (TE; 100 and 200 mg weekly) plus daily oral doses of cyproterone acetate (CPA; 5-100 mg), levonogestrel (LNG; 150-500 micro g), or desogestrel (DSG; 150-300 micro g). Overall, men with sperm concentrations below 0.1 million/ml had significantly lower gonadotropin levels (serum FSH, approximately 0.12 IU/liter; serum LH, approximately 0.05 IU/liter) than oligospermic men (sperm concentrations, 0.1-5 million/ml; serum FSH, 0.23-0.5 IU/liter; serum LH, 0.05-0.56 IU/liter), but the relationship was weak, indicating the possible existence of other determinants. Multivariate logistic regression was used to identify the influence of candidate predictors of spermatogenic effects of the T plus progestin regimens. In the LNG and DSG studies, the marked suppression of serum LH to less than 5% of baseline values (<0.15 IU/liter) was a consistent and highly significant predictor of sperm concentration (reduced to 2-7% that seen at higher LH levels) and the likelihood of its suppression below 1 million/ml (a proposed threshold for contraceptive efficacy). Serum FSH was not a significant independent predictor. The use of DSG and CPA (but not LNG) was a significant independent predictor of sperm suppression, and regimens that contained 200 mg TE weekly caused less spermatogenic suppression than 100 mg TE weekly. These findings suggest that T-progestin contraceptive regimens suppress sperm concentration by gonadotropin-dependent and -independent mechanisms. The suppression of serum LH is a major predictor of the suppression of sperm concentration suppression in the LNG and DSG treatment studies. On the other hand, the greater spermatogenic suppression in regimens containing DSG or CPA suggests that these progestins have additional actions to suppress spermatogenesis via a gonadotropin-independent mechanism(s)