BACKGROUND AND OBJECTIVES: Posttransplant anemia is frequently reported in kidney transplant recipients and is associated with worsened patient survival. Similar to high erythropoiesis-stimulating agent requirements, resistance to endogenous erythropoietin may be associated with worse clinical outcomes in patients with ESRD. We examined the association between serum erythropoietin levels and mortality among kidney transplant recipients. DESIGN, SETTING, PARTICIPANTS, & MEASUREMENTS: We collected sociodemographic, clinical, medical, and transplant history and laboratory data at baseline in 886 prevalent kidney transplant recipients (mean age 51 ± 13 [SD] years, 60% men, 21% diabetics). A solid-phase chemiluminescent immunometric assay was used to measure serum erythropoietin. Cox proportional hazards regression was used to model the association between baseline serum erythropoietin levels and all-cause mortality risk. RESULTS: During the median 39-month follow-up, 99 subjects died. The median serum erythropoietin level was 10.85 U/L and hemoglobin was 137 ± 16 g/L. Mortality rates were significantly higher in patients with higher erythropoietin levels (crude mortality rates in the highest to lowest erythropoietin tertiles were 51.7, 35.5, and 24.0 per 1000 patient-years, respectively [P = 0.008]). In unadjusted and also in adjusted Cox models each SD higher serum erythropoietin level significantly predicted all-cause mortality: HR(1SD increase) 1.22 and 1.28, respectively. In adjusted Cox models each SD higher serum erythropoietin/blood hemoglobin ratio also significantly predicted all-cause mortality: HR(1SD increase) 1.32. Serum erythropoietin predicted mortality in all analyzed subgroups. CONCLUSIONS: In this sample of prevalent kidney transplant recipients, higher serum erythropoietin levels were associated with increased mortality.