Childhood survivors of cancer are prone to a number of adverse sequelae related to the therapeutic interventions undertaken to achieve remission. The endocrine system is frequently affected; hypothalamo-pituitary dysfunction, in particular GH deficiency, is common after cranial irradiation. It is unclear to what extent GH deficiency contributes to the abnormalities observed in adult survivors of childhood cancer, and whether GH replacement reverses these anomalies. We compared 27 GH-deficient survivors of childhood cancer with 27 adult age- and sex-matched controls and went on to replace GH in the patient group to determine whether GH resulted in improvements of the baseline abnormalities. The GH-deficient survivors of childhood cancer had an adverse lipid profile (total cholesterol, 5.4 vs. 4.6 mM, P = 0.004; high-density lipoprotein cholesterol, 1.05 vs. 1.6 mM, P < 0.001; and triglycerides, 1.3 vs. 1.0 mM, P < 0.001) and were osteopenic (lumbar spine z-score, -1.53 vs. -0.31 SD score, P < 0.001; femoral neck z-score, -1.23 vs. -0.27 SD score, P = 0.02); additionally, the female subgroup had an increased percentage body fat (43.6 vs. 32.8%, P = 0.016). In keeping with the selection criterion, quality of life in the patient cohort, relative to the healthy controls, was severely impaired [adult GH-deficiency assessment (AGHDA), 15.5 (range, 8-25) vs. 1 (range, 0-19), P < 0.0001; psychological general well-being schedule, 67.5 (range, 18-86) vs. 89.0 (range, 51-104), P < 0.0001]. After 12 months of GH replacement, small (but significant) improvements were observed in body composition in the male subgroup (waist-hip ratio, 0.871 vs. 0.863, P < 0.05); and in the female cohort, total cholesterol (6.0 vs. 5.2 mM, P = 0.01) and triglyceride (2.1 vs. 1.4 mM, P = 0.01) levels fell. Bone mineral density improved in only one of the four sites studied (ultradistal radius, -1.21 vs. -1.09, P = 0.048) after a median duration of GH therapy of 18 months. Quality of life improved dramatically by 3 months (AGHDA, 15.5 vs. 10.0, P < 0.001), and the improvement was maintained at 12 months (AGHDA, 15.5 vs. 9.0, P < 0.001). Importantly, there was no clinical suggestion of tumor recurrence during the 12 months of GH replacement. The minor improvements observed in body composition, the lipid profile, and bone mineral density in GH-deficient adult survivors of childhood cancer after 12-18 months of physiological GH replacement suggest that GH deficiency may not be the major etiological factor in their pathogenesis; the converse seems to be true for the quality of life status of these individuals. We propose that, as in patients with hypopituitarism caused by pituitary disease, the main indication for GH replacement in GH-deficient survivors of childhood cancer should be severe impairment of quality of life.