GH replacement therapy has been shown to improve abnormalities in body composition, bone mineral density (BMD), lipid profile, and other changes resulting from GH deficiency (GHD) in adults. There is, however, need to determine appropriate dosing in young adults who were treated for GHD as children, to bridge the interval between childhood (in which relatively high doses are used) and older adulthood (in which only lower doses are tolerated). This multicenter, randomized, double-blind, placebo-controlled study compares the safety and efficacy of two doses of GH (25 and 12.5 microg/kg.d) with placebo, maintained for 2 yr, in adults with GHD who were treated as children and were off GH for at least 1 yr (mean, 5.6 yr). The 64 treated subjects were less than 35 yr of age (mean, 23.8 yr) and had maximum serum GH responses, on retesting less than 5 microg/liter (mean, 0.7 micro g/liter). At baseline, 22% had spine BMD below -2 SD, 59% were overweight or obese, and 45% had serum total cholesterol more than 200 mg/dl. A significant dose response was seen for percent increase in spine BMD at 24 months (mean of 1.3%, 3.3%, and 5.2% in the placebo, 12.5-, and 25- microg/kg.d groups, respectively, P = 0.018). Both GH-treated groups had similar changes in body composition at 6 months (decreased fat mass, increased lean mass); however, some gains were subsequently lost in the lower dose group. A significant decrease in low-density lipoprotein cholesterol was seen only in the higher GH dose group. Significant changes were not observed in quality of life and echocardiographic measures. The groups were similar with regard to adverse events and laboratory measurements, except for a higher incidence of edema in the GH-treated groups. We conclude that this dose-response study confirms the benefits of GH-replacement therapy in GHD adults and indicates that, to achieve treatment goals in younger adults, higher doses may be needed than those generally used in older adults.