To evaluate objectively the effects of recombinant human erythropoietin (rHuEPO) administration on nutritional status in stable dialyzed patients, we used urea kinetic modeling (UKM) analysis and dietary protein intake evaluation by dietary assessment. Fifteen patients (9 females, 6 males; mean age 46.9 +/- 15.6 years) dialyzed for 9.4 +/- 6.3 years were studied longitudinally for 18 months, consisting of a control period (6 months) and an rHuEPO treatment period (12 months). Treatment modalities based on 3 weekly sessions were hemodialysis in 12 patients (6 cuprophane, 3 cellulose acetate and 3 highly permeable membranes), hemodiafiltration in 2 patients and postdilutional hemofiltration in 1 patient. Bicarbonate buffered dialysate was used in 9 patients and acetate in 6 patients. Urea kinetic modeling using a single-pool model was performed monthly over 1-3 cycles. rHuEPO was administered intravenously at the end of dialysis according to a two-phase protocol: (1) correction of anemia by stepwise increment of rHuEPO dose, and (2) maintenance dose to keep hemoglobin at 10-11 g/dl. rHuEPO administration corrected anemia in all patients, improving their general clinical condition. Dialysis efficacy was significantly reduced (15%) after the 3rd month of rHuEPO therapy. Clearnces were restored by increasing dialysis time and/or improving dialyzer performances, and adequacy of dialysis was maintained in all patients. During the 12 months of rHuEPO therapy, the protein catabolic rate remained stable at 1.2 g/kg/24 h in spite of an increase in appetite. At the same time, dry body weight increased significantly after 9 months, and the ratio dietary protein intake/protein catabolic rate a gross estimation of nitrogen balance, increased.(ABSTRACT TRUNCATED AT 250 WORDS)