OBJECTIVE: Tiotropium is a new once daily bronchodilator for the treatment of patients with Chronic Obstructive Pulmonary Disease (COPD). Treatment patterns in COPD differ considerably between countries and resource use in one country can not be transferred to other countries without adjustments. The aim of this study was to develop a Markov model to transfer trial-based cost-effectiveness data about tiotropium versus ipratropium, salmeterol and existing therapy from one setting to another. METHODS: Markov states were defined on the basis of COPD severity (moderate A, moderate B and Severe according to GOLD guidelines). Transition probabilities and probabilities of experiencing a severe or non-severe exacerbation were derived from three trial-based cost-effectiveness analyses, which were performed in the Netherlands, Belgium, the USA and in 18 other countries. Resource use was distinguished into maintenance treatment according to disease severity and treatment associated with exacerbations. The model allowed for country-specific input of resource use and unit costs including the probability and duration of hospitalization during an exacerbation. RESULTS: All costs other than the cost of study medication were consistently lower in tiotropium than in the other three treatment arms, mainly because of the 13% to 27% lower probability of experiencing a (severe) exacerbation. Costs in the US were approximately twice as high as the costs in the Netherlands and Belgium. The proportion of total costs due to hospitalization varied between 30% and 60%, depending on the treatment arm and the country. CONCLUSION: The difference between treatment groups observed in the trials could be modeled entirely by the monthly transition probabilities between disease states and the probabilities of experiencing an exacerbation in a given disease state. A Markov model in which resource use depends on disease state and exacerbations, irrespective of treatment group, is well suited to convert trial-based pharmacoeconomic data in COPD to other countries.