OBJECTIVES: To determine the cost-effectiveness of apixaban compared to aspirin to prevent thromboembolic events in patients with atrial fibrillation who are unsuitable or intolerant of warfarin therapy, from an Australian healthcare perspective . METHODS: By extrapolating data from the Apixaban Versus Acetylsalicylic acid to prevent Stroke in Atrial Fibrillation (AVERROSE) trial, a Markov model with yearly cycles was developed to simulate the costs and effects of apixaban compared to aspirin over 10 years. The model comprised five health states: ‘Alive without thromboembolic disease (stroke, myocardial infarction and other systemic embolism) nor major bleeding (MB)’; ‘Alive with thromboembolic disease, but without prior MB’; ‘ Alive without thromboembolic disease, but with prior MB’; ‘Alive with thromboembolic disease and prior MB’; and ‘Dead’. Costs, from an Australian healthcare perspective, were estimated from published sources. The main outcome of interest was incremental cost-effectiveness ratio (ICER) per quality adjusted life year (QALY) saved and per year of life saved (YoLS).  Costs and benefits were discounted at 5.0% per annum. RESULTS: For each patient followed-up over 10 years, the model predicted that compared to aspirin, apixaban would lead to 0.19 YoLS (discounted) and 0.20 QALYs saved (discounted), at a net cost of AUD $5,025 (discounted). This equated to ICERs of AUD $27,090 per YoLS and AUD $25,095 per QALY saved. One way and probabilistic sensitivity analyses indicated the results to be robust. CONCLUSIONS: Compared to aspirin, apixaban is likely to be cost-effective in preventing thromboembolic disease among patients with atrial fibrillation who are intolerant to warfarin.