Background: The aim of this study was to estimate the cost-effectiveness of tiotropium versus salmeterol to inform decision making within the Dutch healthcare setting. Methods: A previously published, validated COPD progression model was updated wih new exacerbation data and adapted to the Dutch setting by including Dutch estimates of healthcare use for COPD maintenance treatment and Dutch unit costs. Exacerbation data from the POET-COPD trial were combined with evidence from earlier tiotropium studies using Bayesian meta-analysis. The model-based analysis was performed using a one- and five-year time horizon. Main health outcomes were the number of exacerbations and quality-adjusted life years (QALYs). Results: One-year costs per patient from the healthcare perspective were v1370 for tiotropium and v1359 for salmeterol; a difference of v11 (95% uncertainty interval (UI): -198-212). The annual number of exacerbations was 0.068 (-0.005-0.140) lower in the tiotropium group. The number of QALYs inthe tiotropium group was 0.011 (-0.019-0.049) higher, resulting in an incremental cost-effectiveness ratio (ICER) of v1015 per QALY. After five years, the difference in exacerbations, QALYs and costs between the tiotropium and salmeterol group were -0.435 (-0.915-0.107), 0.079 (-0.272-0.520) and v-277 (-1586-1074), respectively, indicating that tiotropium was more effective and less costly. Using a societal perspective, tiotropium dominated salmeterol both after one and five years. Conclusion: Tiotropium reduced exacerbations and exacerbation-related costs. After one year the cost per QALY of tiotropium compared with salmeterol was very low, while after five years tiotropium was found to dominate salmeterol.