Bone mineral density (BMD), serum osteocalcin and type I collagen C-telopeptide (ICTP) were assessed in a cohort of 31 (16 males) adults who had received cranial irradiation in childhood as part of their treatment for acute lymphoblastic leukaemia (ALL). Markers of bone turnover were compared with those of 35 age and body mass index (BMI) matched young adults (18 male). Growth hormone status had previously been determined using an insulin tolerance test and arginine stimulation test. Eight patients were classified as severe growth hormone deficiency (group 1), 12 patients as growth hormone insufficient (group 2) and 11 patients as normal (group 3). Vertebral trabecular BMD, lumbar spine and femoral neck integral BMD and forearm cortical bone mineral content (BMC) was measured 17.8 (6.8-28.6) years after cranial irradiation and was expressed as Z (standard deviation) scores. There was a significant reduction in vertebral trabecular BMD (median Z score -1.25, P < 0.001), in lumbar spine integral BMD (median Z score -0.74, P = 0.001), in forearm cortical BMC (median Z score -1.35, P < 0.001), and less so in femoral neck integral BMD (median Z score -0.43, P = 0.03). There was no difference among the growth hormone status groups for the following BMD measurements: vertebral trabecular BMD, lumbar spine integral BMD or femoral neck integral BMD (P = 0.8, P = 0.96 and P = 0.4 respectively). There was only a marginal significant difference for BMD at the wrist between growth hormone status groups (P = 0.04). There was no correlation between the BMD measurements with time since or age at diagnosis and no difference in markers of bone turnover between patients and controls; median serum osteocalcin 13.3 and 12.0 ng ml (P = 0.7), respectively, and for ICTP 5.0 and 4.9 microg L (P = 0.67) respectively. In conclusion, there is a highly significant reduction in BMD in young adults following treatment for ALL in childhood. The reduction in BMD affects both trabecular and cortical bone but did not seem to be related to time since diagnosis, age at diagnosis, or current growth hormone status. Possible explanations include a direct effect of chemotherapy, steroids or both on bone during childhood and hence an effect on the accretion of bone mass. In view of the risk of fractures in patients with osteopenia, adults treated for ALL in childhood may be at an increased risk of bone fractures later in life irrespective of the underlying cause of the osteopenia and thus intervention should be considered.