LHRH agonist analogs induce hypogonadism in man but the mechanism is uncertain. To evaluate this, we treated 13 normal men with 50 micrograms/day D-Trp6,Pro9-NEt LHRH (LHRHA) for periods up to 8 weeks and measured (1) patterns of endogenous gonadotropin and testosterone secretion, (2) gonadal response to exogenous human LH infusions, and (3) gonadotropin and testosterone responses to hourly bolus doses of LHRH. Seven men were evaluated with frequent sampling for 12-h periods every 4 week during treatment with LHRHA. Before treatment, all had three to four spikes of LH in 12 h. On the first day of treatment, the response to LHRHA was tested in four of the men, and there were significant increases in LH, FSH, and testosterone. After 4 weeks, all men had dramatic decreases in mean testosterone levels and blunted or absent gonadotropin responses to acute injection of LHRHA. Mean gonadotropin levels at 4 and 8 weeks were variable; values lower than pretreatment basal levels were found in three men, while unchanged or higher values were found in the remaining four. The pulsatile pattern of LH secretion, characteristic among these men before treatment, was lost during LHRHA therapy. Forty-eight-hour constant infusion of human LH in four other analog-treated men resulted in increases in serum testosterone comparable to those in untreated men. Pulsatile administration of native sequence LHRH to two other men during chronic treatment with LHRHA failed to elicit demonstrable responses in serum gonadotropin or testosterone levels. LHRHA produced a qualitative change in the pattern of LH release from the pituitary. Mean basal LH levels varied during treatment, but the normal pulsatile pattern was diminished. The gonadotropin response to pulsatile administration of LHRH was lost during chronic treatment with LHRHA, but the Leydig cell remained responsive to exogenous human LH. Thus, the locus of action of the analog appears to be at the level of the pituitary in man.