Insulin-like growth factor (IGF) system components appear to be the most important regulators of bone cell function. On the other hand, IGF-1 is shown to be an important regulator for erythropoiesis. The aim of the study was to examine the relationships between IGF system, requirements of erythropoietin, endogeneous erythropoietin levels, bone metabolism assesed by biochemical markers, markers of nutrition such as cholesterol and albumin in recombinant human erythropoietin (rHuEPO)-treated patients maintained on chronic hemodialyses or peritoneal dialyses as well as in kidney transplant recipients. The studies were performed on 79 chronically hemodialyzed patients; 28 of them did not receive rHuEPO, 51 subjects received rHuEPO, 34 patients on continuous ambulatory peritoneal dialysis (CAPD), 16 of them did not receive rHuEPO, 18 were given rHuEPO and 46 kidney allograft recipients. Endogeneous erythropoietin concentration, bone-specific alkaline phosphatase and serum CrossLaps were assayed by ELISA. Intact PTH, osteocalcin, 1,25-(OH)<sub>2</sub> D<sub>3</sub>, 25-OH D<sub>3</sub>, IGF-1, procollagen type I carboxy-terminal extension peptide (PICP) and procollagen type I cross-linked carboxy-terminal telopeptide (ICTP) were studied by RIA, whereas IGFBP-1 and IGFBP-3 concentrations were assayed by IRMA. We found a significantly higher IGF-1 and IGFBP-3 in rHuEPO-treated HD patients when compared to CAPD subjects given rHuEPO as well as to hemodialysis (HD) patients not treated with rHuEPO. IGF-1 was significantly higher in kidney transplant recipients when compared to dialyzed patients without rHuEPO therapy. IGFBP-1 was similar in all groups of patients (including kidney transplant recipients) studied. In CAPD patients not given rHuEPO concentrations of ICTP and PICP were significantly lower when compared to rHuEPO-treated CAPD subjects and HD patients not receiving rHuEPO therapy. Serum CrossLaps in CAPD patients treated with rHuEPO were significantly higher when compared to CAPD subjects without rHuEPO treatment and to kidney transplant recipients. In rHuEPO-treated CAPD subjects IGF-1 and IGFBP-1 correlated positively with serum CrossLaps (r = 0.61, p < 0.05 and r = 0.64, p < 0.05, respectively), whereas in hemodialyzed patients without rHuEPO a significant negative correlation between IGFBP-3 and serum CrossLaps was found (r = –0.69, p < 0.001) as well as between IGFBP-3 and aluminium (r = 0.51, p < 0.05), IGF-1 and ICTP (r = –0.43, p < 0.05). In conclusion, our data indicate a probable functional relationship between IGF system components, erythropoietin treatment in dialyzed patients and bone metabolism in renal replacement therapy in a form of hemodialyses, peritoneal dialyses and kidney transplantation. Dialyzed patients exhibit more pronounced renal osteodystrophy than kidney allograft recipients. IGF system components are influenced by erythropoietin therapy, but are not related to serum erythropoietin levels and rHuEPO requirements.