OBJECTIVE: To investigate whether the changes in lipoproteins following growth hormone (GH) substitution in GH de®cient (GHD) adults are determined by the concomitant changes in body composition and physical ®tness in a controlled long-term study. DESIGN: A randomized, double-blind, placebo-controlled trial with GH (2 IU/m 2 ) or placebo given for 12 months. SUBJECTS: Twenty-seven patients (18 male, 9 female, aged 21±61 y) with adult onset GH de®ciency. Comparisons were made with age-and gender-matched healthy adults. MEASUREMENTS: Serum triglycerides (TG) and lipoproteins, body composition (Dual-Energy X-ray Absorptiometry and computerized tomography), and exercise capacity (V Á O 2 -max measured by bicycle ergometry) were measured at baseline and after 12 months. RESULTS: Baseline values of total cholesterol, low-density lipoprotein (LDL) and serum triglycerides were signi®cantly higher in GHD adults compared to normal subjects (P`0.001, P`0.001, P 0.004, respectively) whereas no difference in high-density lipoprotein (HDL) or lipoprotein (a) (Lp(a)) was found. After one year of GH treatment total cholesterol decreased signi®cantly (P 0.02). Serum LDL decreased after GH and increased after placebo, but the difference in delta values was not signi®cant (P 0.12). Serum HDL and TG concentrations were unchanged. Lp(a) increased but not signi®cantly. Serum total and LDL cholesterol remained signi®cantly elevated after one year of GH treatment. Signi®cant reductions in total and visceral adiposity, and improved exercise capacity were also recorded after GH treatment. In normal subjects, serum total cholesterol and TG correlated positively with age, subcutaneous fat and intraabdominal fat, and negatively with V Á O 2 -max. Serum LDL correlated positively with age. In GHD patients, baseline values of serum TG correlated positively with subcutaneous fat and serum insulin. During treatment, no signi®cant correlations were found between the changes in lipoproteins and in body composition. CONCLUSION: The cholesterol lowering effect of GH is not determined by the concomitant decrease in adiposity, which supports the concept of a direct effect of GH on lipoprotein metabolism.