GH replacement therapy has proved its efficacy and safety in short-term trials and in a few long-term trials with limited number of subjects. In this 1-center study, including 118 consecutive adults (70 men and 48 women; mean age, 49.3 yr; range, 22-74 yr) with adult-onset GH deficiency, the effects of 5 yr of GH replacement on body composition, bone mass, and metabolic indices were determined. The mean initial GH dose was 0.98 mg/d. The dose was gradually lowered, and after 5 yr the mean dose was 0.48 mg/d. The mean IGF-I SD score increased from -1.73 at baseline to 1.66 at study end. A sustained increase in lean body mass and a decrease in body fat were observed. The GH treatment increased total body bone mineral content as well as lumbar (L2-L4) and femur neck bone mineral contents. BMD in lumbar spine (L2-L4) and femur neck were increased and normalized at study end. Total cholesterol and low density lipoprotein cholesterol decreased, and high density lipoprotein cholesterol increased. At 5 yr, serum concentrations of triglycerides and hemoglobin A(1c) were reduced compared with baseline values. The treatment responses in IGF-I SD score, body fat as estimated by four- and five-compartment body composition models, total body protein and nitrogen, and lumbar bone mineral content and BMD were more marked in men than in women. One patient died during the period, four patients discontinued the study due to adverse events, and one dropped out due to lack of compliance. Four patients were lost to follow-up. However, all patients were retained in the statistical analysis according to the intention to treat approach used. In conclusion, 5 yr of GH substitution in GH-deficient adults is safe and well tolerated. The effects on body composition, bone mass, and metabolic indices were sustained. The effects on body composition and low density lipoprotein cholesterol were seen after 1 yr, whereas the effects on bone mass, triglycerides, and hemoglobin A(1c) were first observed after years of treatment.