RATIONALE The efficacy of tiotropium/olodaterol (T/O) compared with tiotropium (Tio) has been demonstrated in a large Phase III clinical program. In this subgroup analysis, we investigated the magnitude of the treatment effect in patients who were not receiving maintenance treatment with long-acting muscarinic antagonists (LAMAs), long-acting β2-agonists (LABAs) or inhaled corticosteroids (ICS) at study entry. METHODS TONADO 1 and 2 (NCT01431274, NCT01431287) and OTEMTO 1 and 2 (NCT01964352, NCT02006732) were randomized, double-blind, parallel-group, Phase III trials in patients with COPD. TONADO 1 and 2 were 52-week studies including patients with post-bronchodilator FEV1 <80% predicted, whereas OTEMTO 1 and 2 were 12-week studies including patients with post-bronchodilator FEV1 ≥30% and <80% predicted. In this post hoc analysis, we examined the treatment differences between T/O 5/5µg and Tio 5µg in trough FEV1 response, St. George’s Respiratory Questionnaire (SGRQ) total score change from baseline and Transition Dyspnea Index (TDI) score at 12 weeks in patients who were not receiving LAMA, LABA or ICS, as monotherapy or combination therapy, at baseline in all four studies. RESULTS The four trials included 1,078 patients in the T/O and Tio arms who were not receiving LAMA, LABA or ICS at trial enrollment. The pooled analysis of TONADO and OTEMTO shows a significantly greater increase in trough FEV1 from baseline with T/O compared with Tio at Week 12 (0.056 L; 95% confidence interval [CI] 0.033 to 0.079; P<0.0001) (Table). There were also significant improvements with T/O versus Tio in SGRQ total score (-1.780; 95% CI -3.126 to -0.434; P=0.0096) and in TDI score (0.409; 95% CI 0.077 to 0.741; P=0.0158) at Week 12. In the pooled analysis, there was a greater chance of being an SGRQ responder (≥4-unit improvement) with T/O treatment (59.6% of patients) than Tio (49.1% of patients; odds ratio 1.5230; 95% CI 1.1844 to 1.9584; P=0.0010), and of being a TDI responder (≥1-unit improvement) with T/O (64.5% of patients) than with Tio (55.7% of patients; odds ratio 1.4758; 95% CI 1.1392 to 1.9118; P=0.0032). CONCLUSIONS Overall, there were greater improvements in lung function, health status and breathlessness with T/O compared with Tio in patients who were not receiving LAMA, LABA or ICS at baseline. This suggests that dual bronchodilation with T/O may be an effective first-line option for treatment-naïve symptomatic patients with COPD.