OBJECTIVE Osteopenia has been reported in adult patients with chronic renal failure (CRF). Only a few studies have been performed in children. The objective of this study was to evaluate bone mineral density (BMD), bone turnover, body composition in children with CRF and to study the effect of GH on these variables. METHODS Two groups were identified: patients with growth retardation who received GH (GH-group) and patients most of whom were not growth retarded who did not receive GH (no-GH-group). After an observation period of 6 months, the patients in the GH-group started GH treatment. Patients were studied every 6 months during 18 months. METHODS Thirty-six prepubertal patients (27 boys and 9 girls), mean age 7.9 years, with CRF participated in the study. The GH-group consisted of 17 patients of whom 14 completed one year treatment. The no-GH-group consisted of 19 patients, of whom 16 were followed for 6 months, 14 for 12 months and 13 for 18 months. METHODS Lumbar spine BMD, total body BMD and body composition were assessed by dual energy X-ray absorptiometry, compared to age-and sex-matched reference values of the same population and expressed as standard deviation scores (SDS). BMD of appendicular bone was measured by quantitative microdensitometry (QMD). Blood samples were obtained to assess bone metabolism and growth factors. RESULTS Baseline mean lumbar spine and total body BMD SDS of all patients were not significantly different from normal. Mean lumbar spine and total body BMD SDS did not change significantly in the GH-group during GH treatment. The change of QMD at the midshaft during the first 6 months of GH treatment was significantly smaller than during the observation period (P < 0.01). Height SDS and biochemical markers of both bone formation and bone resorption increased significantly during GH treatment; 1,25-dihydroxyvitamin D remained stable. Lean tissue mass increased (P < 0.001) and percentage body fat decreased (P < 0.01) during GH treatment. BMD, the biochemical markers of bone turnover which are independent of renal function, and body composition remained stable in the no-GH-group. CONCLUSIONS Mean lumbar spine and total body BMD of children with chronic renal failure did not differ from healthy controls. The lack of a GH-induced increase in 1,25-dihydroxyvitamin D levels, probably due to treatment with alpha-calcidol, might be linked to the absence of a response in BMD during GH treatment in children with chronic renal failure.