The present study tests the postulate that testosterone (Te) stimulates GH secretion, in part, by attenuating autonegative feedback. To this end, 13 healthy men (ages 43-71 yr) received three consecutive weekly im injections of placebo (Pl) (n = 7) or Te (200 mg) (n = 6) in a prospectively randomized, double-blind, parallel-cohort design. An iv pulse of saline or recombinant human (rh)GH (3 micro g/kg.6 min) was infused 2 h before bolus saline or GH-releasing peptide (GHRP)-2 (1 micro g/kg) in the fasting state. Blood was withdrawn every 10 min, GH concentrations were quantitated by chemiluminometry, secretion was determined by deconvolution analysis, and outcomes were compared by ANOVA. After Pl, rhGH suppressed basal, pulsatile, and GHRP-2-stimulated GH secretion by 2.6-, 2.4-, and 2.1-fold, respectively (each P < 0.03), and truncated GHRP-2-stimulated GH secretory bursts (P < 0.005). Compared with Pl, Te: 1) stimulated basal and pulsatile GH secretion by 1.9 and 2.4-fold (P < 0.01 and P < 0.02), respectively; 2) reduced feedback on basal GH secretion (P < 0.01); 3) blunted GHRP-2-stimulation by 1.9-fold (P < 0.01); and 4) facilitated initial recovery of rhGH-suppressed GH concentrations (P < 0.005). The foregoing actions were selective, inasmuch as Te did not relieve autoinhibition of pulsatile GH secretion. In summary, short-term Te supplementation decreases rhGH-imposed negative feedback on basal GH secretion and enhances early escape of GH from autoinhibition. In principle, such actions could potentiate the renewal of high-amplitude pulses of GH in androgen-replete individuals.