The relative roles of FSH and LH in the control of human spermatogenesis are not well established. We previously reported that supraphysiological doses of hCG can stimulate sperm production in gonadotropin-suppressed normal men despite prepubertal FSH levels. To determine whether more nearly physiological levels of human LH (hLH) also can stimulate spermatogenesis when FSH levels are suppressed, we administered hLH to normal men whose endogenous gonadotropin levels and sperm production were suppressed by exogenous testosterone enanthate (T). After a 3-month control period, 11 normal men received 200 mg T, im, weekly to suppress LH and FSH. T administration alone was continued for 3-4 months until 3 successive sperm concentrations (performed twice monthly) revealed azoospermia or severe oligospermia (sperm concentrations, less than 4 million/ml). Then, while continuing T, 4 of the 11 men (experimental subjects) simultaneously received 1100 IU hLH, sc, daily for 4-6 months to replace LH activity, leaving FSH activity suppressed. The effect on sperm production of the selective FSH deficiency produced by hLH plus T administration was determined. The remaining 7 men (control subjects) continued to receive T alone at the same dosage, without gonadotropin replacement, for an additional 6 months. In the four experimental subjects, sperm concentrations increased significantly from 0.7 +/- 0.7 million/ml (mean +/- SEM) during T treatment alone to 19 +/- 4 million/ml during hLH plus T administration (P less than 0.001). However, none of the men achieved sperm concentrations consistently in their own pretreatment range. Sperm motilities and morphologies were normal in all four subjects by the end of hLH plus T administration. In contrast, sperm concentrations in the seven control subjects remained suppressed (less than 3 million/ml) throughout the entire period of prolonged T administration alone. Serum LH bioactivity, determined monthly by in vitro mouse Leydig cell bioassay in all four experimental subjects, was markedly suppressed during T administration alone (120 +/- 10 ng/ml) compared to that during the control period (390 +/- 20 ng/ml; P less than 0.001). With the addition of hLH to T, LH bioactivity returned to control levels (400 +/- 40 ng/ml; P = NS compared to control value). Serum FSH levels determined monthly by RIA were reduced from 98 +/- 12 ng/ml during the control period to undetectable levels (less than 25 ng/ml) during the T alone and the hLH plus T periods (P less than 0.01).(ABSTRACT TRUNCATED AT 400 WORDS)