Information about the site(s) of action as well as the age-dependent effects of sex steroids on gonadotropin- releasing hormone and gonadotropin secretion during human puberty is limited. To begin to address these questions, we evaluated the effects of a depot preparation of testosterone (testosterone enanthate) on gonadotropin secretion and pituitary responses to synthetic GnRH in 10,early to mid-pubertal boys who had either isolated GH deficiency (n-2) or delayed adolescent maturation (n-8). Chronological and bone age ranges were 13 1/12—16 1/12 and 11-14 yr, respectively. Frequent blood withdrawal studies (every 20 min for 20 consecutive h) were performed in the Clinical Research Center over two consecutive weekends. Following each study, gonadotropin responses to GnRH (0.25 µg/kg iv bolus) were determined. During the initial study, all boys showed a sleep-entrained increase in luteinizing hormone (LH) and testosterone (T) secretion; mean nocturnal concentrations of LH and T were 2.3-fold greater than daytime values. At the end of the first study, testosterone enanthate was given im (0, 25, 50, or 75 mg/m2). Six days later, mean plasma T concentrations were in the pubertal to mid adult male range and were constant throughout the day: 25 mg/mm2, 3.7 ± 0.4 (SE) ng/ml; 50 mg/m2, 4.6 ± 0.2 ng/ml; and 75 mg/mm2, 6.7 ± 0.4 ng/ml. T treatment had no effect on pituitary responses to GnRH: mean LH increment was 8.5 mlU/ml before and 10.0 mlU/ ml after T treatment. Plasma LH and follicle-stimulating hormone were dramatically suppressed by T; paired analyses, however, revealed the persistence of slightly greater LH values during sleep in four of the nine treated boys. During the control study, LH pulse frequency averaged 2.8 ± 0.3 pulses/6 h during the day and 3.8 ± 0.3 pulses/6 h during sleep. In boys who received 50 or 75 mg/mm2of testosterone enanthate, gonadotropin secretion was profoundly suppressed and LH pulse frequency could not be accurately assessed. However, LH pulse frequency in three boys treated with 25 mg/mm2was not different than their control values: control day, 2.6 ± 0.6; control night, 4.1 ± 0.6; treatment day, 2.1 ± 0.5; and treatment night, 3.9 ± 0.3 pulses/6 h. These results imply that reduction of GnRH secretion is the principal feedback mechanism of testicular steroids in pubertal boys.