OBJECTIVE: Currently imatinib resistant or intolerant CML patients have minimally effective therapies available. Dasatinib binds to the protein Bcr-Abl; it binds also to active and inactive forms of protein, while imatinib binds only to the inactive forms. Therefore, mutations that affect the active form can lead to resistance to imatinib. A Markov model was built to evaluate the long-term cost-effectiveness of dasatinib in the treatment of adult CML patients, after resistance or intolerance to imatinib. METHODS: The model consists of an initial within-trial period in which best response rates observed from the clinical trials are used. Response was defined as best response of complete hematologic response (CHR), minor cytogenetic response (CyR), minimal CyR, partial CyR, and complete CyR. The model simulates patients moving between health states using progression probabilities derived from the literature and BMS clinical trials. The time horizon was the lifetime of patients in the cohort, allowing evaluation of life expectancy and lifetime costs. Brazilian costs and health resource estimates were applied to the treatment of the different phases of CML. RESULTS: For CML patients in CP dasatinib provided 0.66 QALYs per patient and the ICER was R$80,000 with an additional life expectancy of 0.98 years. In the case of AP dasatinib provided an additional life expectancy of 3.48 years with a ICER of R$91,000. And in the BP dasatinib provided an additional life expectancy of 1.91 years with a ICER of R$123,000 CONCLUSION: The CE analysis showed that dasatinib is more cost-effective in the resistant or intolerant patients than imatinib in the three phases of CML with increased life expectancy with quality. Though there is an incremental cost associated to the treatment with dasatinib, the cost is related to longer life expectancy and therefore expenditure of more resources.