OBJECTIVES To assess real-world effectiveness of BUD/FM DPI versus tiotropium DPI in COPD patients METHODS Data from the Quebec health care databases were used to construct a matched cohort of COPD patients aged d40 years newly treated with BUD/FM DPI or tiotropium DPI from 2003-2007. Patients were matched on age, sex, COPD exacerbations (short-course prescription of oral corticosteroids [OCSs], emergency department [ED] visits or hospitalizations for COPD; d2 events occurring within 15 days counted as 1 exacerbation), and use of short-acting inhaled ≤]2-agonists (SABAs) and ipratropium in the year before therapy began. The number of exacerbations, ED visits, and hospitalizations for COPD; claims for OCS prescriptions; and the average weekly doses (dose = 2 inhalations) of SABAs and ipratropium were compared for BUD/FM DPI versus tiotropium DPI users for a 1-year post-therapy period. Poisson and linear regression models were used to produce adjusted rate ratios (RR) and mean differences (MD). RESULTS Of 981 BUD/FM DPI and 981 tiotropium DPI users in the cohort, 78% were aged „d 65 years and 53% were men. No significant differences were seen for COPD exacerbations (RR=0.94; 95% CI, 0.77-1.15), ED visits for COPD (RR=0.80; 95% CI, 0.54-1.20), and claims for OCS prescriptions (RR=0.93; 95% CI, 0.72-1.21) between BUD/FM DPI and tiotropium DPI users in the year after the start of therapy. However, BUD/FM DPI users had significantly fewer hospitalizations for COPD (RR=0.65; 95% CI, 0.44-0.97), used less SABAs (MD=-0.48; 95% CI, -0.67 to -0.28), and used more ipratropium (MD=0.35; 95% CI, 0.21-0.50). CONCLUSIONS These findings showed that patients using BUD/FM DPI were significantly less likely to have COPD exacerbations leading to a hospitalization, but as likely as tiotropium DPI users to require emergency care or OCS therapy in the year after initiation of therapy.