Background: A 6-week, multicenter, randomized, double-blind, parallel-group study was conducted in patients with COPD to compare lung function improvements of tiotropium, is mu g qd, plus formoterol, 12 mu g bid, to salmeterol, 50 mu g bid, plus fluticasone, 500 mu g bid. Methods: Following a screening visit, subjects entered a run-in period in which they received regular ipratropium. At randomization, patients were assigned to either tiotropium plus formoterol or salmeterol plus fluticasone. After 6 weeks of treatment, a 12-h lung function profile was obtained. The coprimary end points were FEV1 area under the curve for the time period 0 to 12 h (AUC(0-12)) and peak FEV1. Results: A total of 729 patients were screened, and 605 patients were randomized and treated. A total of 592 patients (baseline FEV1, 1.32 +/- t 0.43 L/min [+/- SD] were included in the analysis. After 6 weeks, the 12-h lung function profiles in the group receiving tiotropium plus formoterol were superior to those in the group receiving salmeterol plus fluticasone (mean difference in FEV1 AUC(0-12), 78 mL [p = 0.0006]; mean difference in FVC AUC(0-12), 173 mL, p < 0.0001). Also, peak responses were in favor of tiotropium plus formoterol (difference in peak FEV1 103 mL [p < 0.0001]; difference in peak FVC, 214 mL [p < 0.0001]), as were FEV1 and FVC at each individual time point after dose (p < 0.05). Predose FVC was significantly higher with the bronchodilator combination, while predose FEV1 and rescue medication use did not differ significantly between groups. Both treatments were well tolerated. Conclusions: Tiotropium plus formoterol was superior in lung function over the day compared to salmeterol plus fluticasone in patients with moderate COPD. Long-term studies in patients with severe COPD are warranted to assess the relative efficacy of different treatment combinations. Trial registration: Clinicaltrials.gov Identifier: NCT00239421.