Objective: To investigate the effects of long-term GH in GH-deficient adults, as predicted by IGF-I levels. Methods: Patients received GH, 5 mg/kg per day for 1 month and 10 mg/kg per day for another 12-30 months. Changes in body composition, cardiac structure/function, serum lipids and quality of life were measured. Results: There was a significant increase in lean body mass (LBM) (2.21 kg; P , 0.0001) after 6 months, which was sustained throughout treatment. A larger increase occurred in males than females (2.97 vs 1.19 kg; P , 0.0001). Total fat mass was reduced (2.56 kg; P , 0.0001 (3.26 kg males, 1.63 kg females)). Responsiveness to GH varied greatly, but LBM changes correlated with IGF-I changes (P , 0.004). Furthermore, thinner patients experienced greater and progressive LBM increases. There was an increase in ejection fraction (3.85^9.95%; P ¼ 0.0002) after 6 months, sustained to 18 months. These cardiac effects were equal for males and females, and did not correlate with IGF-I levels. Serum low-density lipoprotein/high-density lipoprotein ratios decreased within 6 months, and were sustained thereafter. Quality of life improved significantly after 6 months, an effect that was sustained/enhanced as treatment continued. No major adverse events were identified. Conclusions: Improved body composition is both reflected by IGF-I changes and predicted inversely by baseline adiposity. Other effects of GH replacement on cardiac function, dyslipidaemia and quality of life, however, do not correlate with circulating IGF-I concentrations. Our findings validate the importance of sustained GH therapy, but caution on the interpretation of IGF-I levels in monitoring the longterm effects of GH treatment.