OBJECTIVE: The serum concentration of the high-af®nity growth hormone-binding protein (GHBP) is increased in obesity but the mechanisms are poorly understood. This study assessed the physiological mechanisms involved in the regulation of GHBP in adiposity. SUBJECTS AND MEASUREMENTS: We tested a number of obesity speci®c parameters for their association with GHBP. In this study, 199 normal or overweight children and adolescents (101 boys, 98 girls, aged (mean AE s.d.): 13.7 AE 2.3 y) underwent an anthropometric evaluation (circumference measurements and bioimpedance analysis) combined with blood withdrawal for the measurement of insulin-like growth factor-I (IGF-I), insulin, leptin and GHBP (by speci®c RIA), uric acid, triglycerides and cholesterol. RESULTS: By linear regression analysis GHBP correlated signi®cantly (P`0.001) with percent body fat mass (r 0X71), waist (r 0X73) and hip (r 0.69) circumference, weight (r 0.61), waist hip ratio (WHR) (r 0.54), as well as with the serum concentrations of leptin (r 0.64), uric acid (r 0.54), insulin (r 0.45), LDL-cholesterol (r 0X43), cholesterol (r 0.33), LDLaHDL ratio (r 0.47), triglycerides (r 0.30) and with height standard deviations scores (SDS) (r 0.23). Age, gender and pubertal stage had no impact on GHBP. In a multiple regression analysis containing age and gender, as well as the anthropometric variables, percent fat mass and waist circumference, as independent variables, associations between GHBP and leptin (P`0.001), cholesterol (P`0.01), LDL-cholesterol (P 0.01), LDLaHDL ratio (P 0.02), triglycerides (P 0.01) remained signi®cant. In a ®nal model using the stepwise analysis involving age, gender and all the independent predictors of GHBP, waist circumference (P`0.001) accounted for 49.5% of the 60.0% total variability in GHBP, while the implication of leptin (P`0.001), age (P`0.01) and cholesterol(P`0.05) increased the predicted variability for 7.5%, 1.9%, and 1.0%, respectively. Serum GHBP was signi®cantly reduced in a subgroup of 104 overweight or obese patients during a diet-induced weight loss programme, the coef®cient of correlation between GHBP and leptin after (r 0.45, P`0.001) and before weight reduction (r 0.41, P 0.001) were comparable. CONCLUSION: Waist circumference, an indicator of abdominal body fat mass, is a major determinant of GHBP levels during childhood, while leptin may be one candidate for a signal linking adipocytes to the growth hormone receptor related GHBP release. Additionally, elevated serum levels of GHBP may re¯ect metabolic disturbances of adiposity.