Extended Abstract Low birth weight has been associated with impaired insulin sensitivity, type 2 diabetes mellitus, hypertension and cardiovascular disease in later life. GH therapy is known to increase fasting and postprandial insulin levels. For this reason concern has been expressed regarding the possible detrimental effects of GH therapy in children born small for gestational age (SGA). To assess the effects of GH therapy on body composition, carbohydrate metabolism and final height in short SGA children, 165 prepubertal short children born SGA were enrolled in either a multicentre, double-blind, randomized, dose-response GH trial (n = 75) or in a GH controlled trial (n = 90). The inclusion criteria were: (1) birth length standard deviation score (SDS) below –2; (2) age 3–8 years; (3) height SDS below –2. The children’s mean (SD) age was 7.3 (2.1) years (GH dose-response trial) and 6.0 (1.5) years (GH controlled trial), birth length SDS was –3.6 and height SDS was –3.0 (0.7). In the GH dose-response trial, children were randomly assigned to either 1 mg GH/m<sup>2</sup> per day (group A, n = 41) or 2 mg GH/m<sup>2</sup> per day (group B, n = 38) (≈0.033 or 0.067 mg/kg per day, respectively). In the GH controlled trial, children were randomly assigned to 1 mg GH/m<sup>2</sup> per day (n = 60) or served as controls (n = 30). Subjects underwent standard oral glucose tolerance tests and measurement of body mass index, systolic and diastolic blood pressure and serum lipids at baseline and after 1 and 6 years of GH therapy and again 6 months after discontinuation of GH. Body composition was measured by dual energy x-ray absorptiometry at baseline and again after 3 years in the GH controlled trial. Mean (SD) final height SDS was not significantly different between the two GH dosage groups: –1.2 (0.7) in group A and –0.8 (0.7) in group B. At the start of GH therapy, 8% of children had impaired glucose tolerance (IGT). Systolic blood pressure was significantly higher in comparison with healthy peers. GH therapy induced considerably higher fasting and glucose-stimulated insulin levels after 1 and 6 years, regardless of GH dosage. After 6 years, 4% of children had IGT. Six months after discontinuation of GH, glucose levels remained normal, whereas fasting and glucose-stimulated insulin returned to levels comparable to those of healthy peers. None of the children developed diabetes. During 6 years of GH therapy both systolic and diastolic blood pressure decreased significantly and remained so after discontinuation of GH therapy. At baseline all children had reduced bone mineral content and lean body mass. Fat mass was not significantly lower than normal. Treatment with 1 mg GH/m<sup>2</sup> per day resulted in a significant increase in (and normalization of) bone mineral content and lean body mass in comparison with untreated short SGA controls. Fat mass decreased during the first year of GH but returned to values comparable to those at baseline in the following 2 years of GH therapy. We found that long-term, continuous GH therapy in short children born SGA leads to a normalization of height during childhood and to a normal final height in most children, regardless of GH dosage. Only very short or relatively older children may need a dosage of 2 mg GH/m<sup>2</sup> per day. Long-term GH therapy had no adverse effects on glucose levels and serum lipids and had a positive effect on blood pressure, even with GH dosages of up to 2 mg/m<sup>2</sup> per day. However, as has been reported in other patient groups, GH induced higher fasting and glucose-stimulated insulin levels, indicating insulin resistance. After discontinuation of GH serum insulin levels returned to normal age-reference levels. Short SGA children have a reduction in bone mineral content and lean body mass when compared with healthy controls, which significantly improved (normalized) with GH therapy at a dose of 1 mg/m<sup>2</sup> per day.