OBJECTIVES: This study compared drug utilization patterns and associated costs as well as dosing trends over time in patients with chronic kidney disease (CKD) not on dialysis receiving epoetin alfa (EPO) or darbepoetin alfa (DARB) through a pharmacy benefit.  METHODS: An analysis of pharmacy claims between July 2002 and March 2009 from the Ingenix IMPACT database was conducted. Patients ≥18 years, newly initiated on erythropoiesis-stimulating agents (ESAs), with ≥1 claim for CKD, and ≥1 ESA pharmacy claim were included. Patients diagnosed with cancer, receiving chemotherapy or dialysis, or receiving both agents were excluded. Mean cumulative ESA dose was used to calculate drug cost (using 10/2009 wholesale acquisition cost) and dose ratio (Units EPO: mcg DARB). Average weighted weekly ESA dose was calculated during the treatment episode to assess ESA utilization trends over time. Multivariate analysis was also conducted. RESULTS: A total of 4,202 ESA-treated patients were identified (EPO 3,111; DARB 1,091). EPO patients were slightly older (60.1 vs. 57.0, P<.001) with a higher Charlson Comorbidity Index (1.68 vs 1.44, P<.001). Mean [SD] cumulative dose was 219,060 [236,830] Units for EPO and 818 [956] mcg for DARB, resulting in a dose ratio of 268:1 (Units EPO: mcg DARB). The corresponding drug cost was 28% higher for DARB than EPO ($3163 vs. $4039, P<.001). From 2002 to 2009, a decreasing trend was observed in semi-annual mean weekly doses of ESAs [EPO: 17,053 to 13,674 Units (25% decrease); DARB: 63 to 51 mcg (20% decrease)]. After adjusting for potential confounding factors, the DARB cost premium and the decreasing weekly ESA dosing trend over time remained significant. CONCLUSIONS: This study of CKD patients not on dialysis receiving ESAs through pharmacy benefits reported significantly higher drug costs in the DARB group compared with the EPO group and a decreasing ESA dosing trend during the 7-year study period.