Background: Some HIV-infected patients develop fat maldistribution with visceral adipose tissue (VAT) accumulation and metabolic abnormalities. No medical treatment is approved by the US Food and Drug Administration to reduce VAT. Methods: In this double-blind trial, 245 HIV-infected patients with excess VAT were randomized to receive placebo (PL), recombinant human growth hormone (r-hGH) at a dose of 4 mg daily (DD) of 4 mg on alternate days (AD) for 12 weeks. For weeks 12 to 24, DD patients were rerandomized to PL (DD-PL) or AD (DD-AD), AD patients continued on AD (AD-AD), and PL patients were switched to DD (PL-DD). Results: From baseline to week 12, VAT decreased significantly compared with PL in DD (-8.6%, P < 0.001) but not in AD (-4.2%, P = 0.052). Trunk-to-limb fat ratio decreased significantly in both (P < 0.001) compared with PL, as did total cholesterol and non-high-density lipoprotein (HDL) cholesterol (-4.5% and -7.5% in DD, -4.3% and -6.2% in AD). At week 24, all groups displayed significant (P < 0.05) reductions in VAT (-5.3% to -9.5%) and trunk fat (-7.8% to -22.8%). DD-AD and AD-AD also displayed significant (P < 0.05) reductions in non-HDL cholesterol. Conclusions: These results suggest that r-hGH dosed at 4 mg daily for 12 weeks decreases VAT and cholesterol concentrations in HIV-infected patients with excess VAT. The optimal regimen to sustain these effects awaits determination.