Recombinant human erythropoietin (r-HuEPO) has been demonstrated to be efficacious in raising and maintaining hemoglobin levels for hemodialysis patients and those in the predialysis stage of chronic renal failure. However, acceptance of this therapy by governments and other third-party payers is not assured by this clinical success alone. Studies that compare the social and economic value of r-HuEPO to its cost may play an important role in determining financing and use of this therapy. This report discusses the socioeconomic methods that might be used to evaluate new therapies and examines the potential costs and health benefits of r-HuEPO. The vast majority of hemodialysis patients are anemic. Currently, the primary treatment is transfusion therapy for severe asymptomatic anemia or symptomatic anemia (eg, angina or extreme fatigue). Clinical trials have demonstrated that the use of r-HuEPO can virtually eliminate transfusions and ameliorate symptoms typically associated with both organ hypoxia (fatigue and angina) and other disorders (anorexia, insomnia, sexual dysfunction) experienced by chronic renal failure patients. As a result, there are several socioeconomic consequences that can be evaluated. A multicenter clinical trial in Canada has already demonstrated statistically significant improvements in many dimensions of quality of life. In terms of monetary consequences, the direct costs of transfusions will be reduced or eliminated, as will the costs of transfusion-related illness. Third, graft survival for patients who are sensitized by transfusion and later transplanted will likely improve. Other potential benefits have been hypothesized, but not yet examined: that hospitalizations due to angina, myocardial infarction, seizure, and infection might be reduced; and that labor force participation may increase. There will also be a significant cost to using r-HuEPO. Preliminary price information suggests average annual per patient cost of $6,000 in the United States and $8,900 in Canada (Canadian dollars). Moreover, clinical trial results suggested that additional expenditures for hypertensive medications and access-related problems are likely. Socioeconomic studies will eventually provide information concerning the relative magnitude of these costs and consequences. Preliminary estimates suggest that the elimination of transfusions alone could offset from 25% to 50% of costs. However, such dollar estimates will only represent a part of r-HuEPOs beneficial consequences. Physicians, patients, and third-partypayers will have to compare any net costs of this therapy to the value of the quality-of-life improvements that will result from its use.