OBJECTIVES: The objective of this cost-utility analysis was to compare the costs and effectiveness of dabigatran (DAB) with adjusted dose warfarin (WAR), aspirin (ASP) or no treatment (UNT) in patients with Non-Valvular Atrial Fibrillation (NVAF), from the perspective of the third party payer in the Australian public healthcare system. METHODS: A Markov cohort model was constructed based on the pivotal RE-LY clinical trial and indirect comparisons with aspirin and no treatment. The model calculated the incremental cost per QALY of different dabigatran doses (150 mg BID or 110 mg BID) relative to adjusted-dose warfarin, aspirin or no treatment. The model estimated the number of ischaemic strokes (IS), haemorrhagic strokes (HS), systemic embolic events (SEE), intracranial haemorrhages (ICH), transient ischaemic attacks (TIA), extra-cranial haemorrhages (ECH), minor bleeds (MB) and myocardial infarctions (AMI) associated with the respective treatments. The key consequences of the clinical events were costs, disability and/or reduction in quality of life, and death. The costs, morbidity and mortality of IS and HS in Australia were estimated from a patient audit of 3,307 strokes in Australia. RESULTS: The dabigatran treatment groups were associated with greater life years and QALYs compared with all the other treatment groups. These gains were primarily driven by a lower incidence of IS, SEE, TIA, ICH and HS in most comparisons. Incremental cost per QALY ratios were calculated by comparing the expected utilisation of the dabigatran doses (50% of patients using each dose) with current utilisation of adjusted-dose warfarin (40%), aspirin (40%) and no treatment (20%). The incremental cost per QALY of dabigatran was $10,028. Sensitivity and subgroup analyses consistently demonstrated the cost-effectiveness of dabigatran. CONCLUSIONS: Treating patients with dabigatran represents a cost-effective treatment for preventing strokes in patients with NVAF in Australia.