This study was designed to explore whether low doses of recombinant human (rh)GH affect lipid, glucose, or protein metabolism in men with visceral obesity. Four different studies were performed in six, otherwise healthy, obese men (age, 42 +/- 3; body mass index, 33 +/- 1 kg/m(2); waist circumference, 111 +/- 3 cm; mean +/- SEM). Lipid, glucose, and protein kinetics was estimated by infusing stable isotopes (glycerol, glucose, leucine) in the basal state and after 1 wk of treatment with sc bedtime injections of either placebo, 2.5 (GH2.5), or 3.3 (GH3.3) microg rhGH/kg body weight per day. When compared with baseline, placebo had no effect on lipid, glucose, or protein fluxes. In contrast, GH2.5 and GH3.3 increased lipolysis by approximately 25% (P < 0.04) without changing glucose and protein turnover rates. The two rhGH treatments increased within the normal range serum IGF-I (by approximately 30%; P < 0.01), whereas they augmented insulin secretion (P < 0.04) in a dose-dependent manner (GH2.5 by 19%, GH3.3 by 37%). C-peptide secretion was increased (P = 0.01) only by GH3.3 (by 28%). In conclusion, 1 wk of treatment with low doses of rhGH is sufficient to increase lipolysis in visceral obese men, but it does not modify glucose and protein turnover rates. The results of this study provide the rationale to design clinical trials using low doses of rhGH to attempt to reduce fat mass.