OBJECTIVES: To compare retrospective data of erythropoiesis-stimulating agent (ESA) dosing patterns and costs among hospital inpatients with chronic kidney disease (CKD) not on dialysis. METHODS: Electronic records from the Premier Hospital Database (2006Q1-2011Q1) were used to identify inpatients that had a diagnosis for CKD and had received epoetin alfa (EPO) or darbepoetin alfa (DARB). Exclusion criteria were: <18 years, receipt of renal dialysis, a diagnosis of cancer or myelodysplastic syndrome, receipt of chemotherapy, or receipt of both ESAs. The observation period consisted of the hospital inpatient stay. The dose only ratio (Units EPO: mcg DARB) was calculated using the mean cumulative dose of EPO and DARB. Treatment costs for both ESA groups were determined using cumulative dose and published July 2011 wholesale acquisition costs. RESULTS: A total of 135,432 inpatient stays (EPO: 104,195; DARB: 31,237) were identified. The EPO group was slightly younger than the DARB group (69.5 vs. 70.0 years, respectively; P<0.001), and had a higher proportion of females (52.1% vs. 51.2%, respectively; P=0.022). Length of stay (LOS) was similar with mean LOS of EPO=9.3 vs. DARB=9.2 days, (P=0.084). The mean cumulative dose was EPO 35,355 Units and DARB 128 mcg, which resulted in a dose only ratio (Units EPO : mcg DARB) of 276:1. Corresponding mean ESA treatment costs were higher for DARB than for EPO (EPO: $573 vs. DARB: $718; P<0.001). CONCLUSIONS: In this analysis of non-dialysis CKD inpatient records, a dose only ratio (Units EPO: mcg DARB) of 276:1 was observed. Mean ESA treatment costs were found to be approximately 25% higher for the DARB group versus EPO group despite similar length of stay between the groups.