Objectives: Management with vitamin K antagonists (VKAs) has been an effective and cost-effective strategy for stroke prevention in atrial fibrillation (AF) but is associated with shortcomings. Novel oral anticoagulants (NOACs) were developed with the aims of no monitoring requirement and improved effectiveness and safety profiles. Economic evaluations require the comparison of all relevant options. However, there are no randomized controlled trials (RCTs) directly comparing these agents. In such cases, indirect treatment comparison (ITC) can be used to synthesize indirect comparative evidence. Through ITC-based evidence synthesis the cost-effectiveness of all available NOACs for stroke prevention in AF patients may be evaluated. Methods: ITC models were based on RCTs data comparing dabigratran, rivaroxaban, or apixaban with VKA treatment. Relative effectiveness was estimated for stroke/systemic embolism, intracranial hemorrhage, myocardial infarction, extracranial hemorrhage, and minor bleeding. A Markov model was developed using ITC-synthesized evidence with VKA as the baseline. Health utilities were collected from international sources whereas costs and mortality data were extracted from Dutch sources. Univariate and probabilistic sensitivity analyses (PSA) were conducted on the base-case incremental cost-effectiveness ratio (ICER). Results: The ICERs for dabigatran, apixaban, and rivaroxaban compared to VKA were € 12,146/QALY, € 12,488/QALY, and € 24,124/QALY, respectively. Sensitivity analysis using the upper and lower limits of the 95% confidence interval for absolute stroke risk with VKA treatment resulted in ICERs that varied drastically from dominance for VKA to being dominated by all NOACs. This is likely due to the large uncertainty observed between the baseline risk profiles of the VKA arms in the three RCTs. The options with the highest probabilities of cost-effectiveness in PSA were VKA at thresholds under € 13,000/QALY and dabigatran or apixaban at thresholds above this mark. Conclusions: ...