OBJECTIVES: To compare real-world healthcare resource utilization (HCRU) and healthcare costs among patients initiated roflumilast compared to those initiated other chronic obstructive pulmonary disease (COPD) medications. METHODS: A retrospective database analysis was performed using Lifelink^TM Health Plan Claims Database. We included patients with a diagnosis of COPD who initiated roflumilast (roflumilast group) or any other ≥3 COPD maintenance drugs (non-roflumilast group) between May 1, 2011 and December 31, 2012. Patients must have been enrolled for 12 months prior to (baseline period) and 3 months after (follow-up period) the initiation date, be ≥40 years old, not corticosteroid-dependent, and without asthma diagnosis at baseline. Office and emergency department (ED) visits, hospitalizations, and total healthcare costs were compared between study groups using nonparametric Wilcoxon rank-sum test. Difference-in-difference (DID) models were used to compare the change from baseline in hospitalizations and costs between groups while adjusting for baseline differences. RESULTS: During the baseline period, patients in the roflumilast group (N=710) had significantly more office visits, ED visits, hospitalizations, and greater total costs than patients in the non-roflumilast group (N=13,501). The changes from baseline were: monthly frequency of office visits (0.081 in roflumilast vs 0.122 in non-roflumilast group; p=0.01), ED visits (0.009 vs 0.007; p=0.80), hospital admissions (0.002 vs 0.005; p=0.02), and monthly costs ($432 vs $522; p=0.34). After controlling for key covariates using DID models, roflumilast was associated with numerically lower hospital admissions (β=-0.003, p=0.57) and total costs (β=-116, p=0.62). CONCLUSIONS: Smaller increase in office visits, hospitalizations and total costs during follow-up from baseline was observed in roflumilast vs non-roflumilast group. A longer follow-up time may be needed to assess potential improvement in these economic outcomes.