BACKGROUND: This research was conducted during a review of the manufacturer’s submission (MS) to the NICE Single Technology Appraisal programme for the oral direct factor Xa inhibitor, rivaroxaban. OBJECTIVES: New anti-thrombotic drugs are available for prevention of stroke in patients with NVAF but evidence on their clinical effectiveness compared with existing treatments is limited. This research compared the clinical effectiveness of rivaroxaban, dabigatran etexilate (dabigatran), aspirin and adjusted standard dose warfarin (warfarin) in people with NVAF. METHODS: Randomised controlled trials (RCTs) for inclusion were identified using the MS for rivaroxaban, and 2 similar reports for dabigatran; inclusion was validated using published systematic reviews. RCTs were assessed for comparability based on patient population, disease severity, and treatments received. A Bayesian MTC was conducted, and fixed and random effects models were explored. Consistency was assessed via pair-wise meta-analysis for each treatment versus warfarin. Odds ratio (OR) was chosen as the summary statistic. RESULTS: The network of 8 RCTs formed a “radiating star”. The fixed effects model was the best-fitting model. There was reasonable agreement between the number of unconstrained data points, residual deviance and pair-wise results, suggesting a coherent network. Statistically significant results compared with warfarin were: reduction in ischaemic stroke with dabigatran (OR 0.78; 95% Credible Interval [95%CrI]: 0.60–1.00); reduction in SE with rivaroxaban (OR 0.24; 95%CrI: 0.07–0.54); reduction in minor extracranial bleeds with dabigatran (OR 0.88; 95%CrI: 0.82–0.96) and aspirin (OR 0.57; 95%CrI: 0.33–0.91); reduction in intracranial bleeds with dabigatran (OR 0.41; 95%CrI: 0.27–0.60) and rivaroxaban (OR 0.66; 95%CrI: 0.46–0.92); increase in myocardial infarction with dabigatran (OR 1.43; 95%CrI: 1.02–1.97); and increase in discontinuations with dabigatran (OR 1.36; 95%CrI: 1.24–1.48). CONCLUSIONS: This research suggests dabigatran and rivaroxaban may offer different clinical benefits and harms in patients with NVAF compared with warfarin.