The cost-effectiveness of epoetin alfa therapy for anemia in 20 patients with end-stage renal disease was retrospectively studied. Ten patients on continuous ambulatory peritoneal dialysis (CAPD) were given subcutaneous epoetin alfa as part of a multicenter, protocol-controlled study of the efficacy of epoetin alfa. Ten patients on in-center hemodialysis were given intravenous epoetin alfa as part of their routine clinical care. Change in hematocrit was used as the measure of effectiveness of epoetin alfa. Medication, laboratory, and transfusion costs were monitored for the six months preceding the initiation of epoetin alfa and the first six months of treatment. The cost of therapy increased for all patients by an average of $2722 +/- 1118; transfusion costs decreased, whereas medication and laboratory costs increased. Laboratory costs were significantly greater in CAPD patients than in hemodialysis patients during epoetin alfa therapy; no significant differences in medication costs or transfusion costs were noted between the groups. The mean increase in hematocrit for all patients was 7.4 volume percent. Following the initial change in hematocrit, further therapeutic response did not appear to be determined by increasing expenditures. Epoetin alfa was shown to be effective in treating anemia in patients with end-stage renal disease, but it was associated with higher costs of therapy.