OBJECTIVES: To compare the effectiveness and safety of oral anticoagulants (OACs) -NOAC and VKA- in patients with atrial fibrillation (AF) in real-world clinical practice. METHODS: Retrospective population-based cohort study. All new users of OAC from Nov2011 to Feb2014 with AF were included (N=21841). Data were obtained by linking diverse electronic databases of the Valencia Region, Spain. Incidence rates were estimated. Cox proportional hazards models adjusted for sociodemographic, clinical, healthcare use data, and propensity scores assessed the risk of several outcomes on effectiveness and safety of each NOAC compared with acenocumarol. Apixaban and warfarin patients were excluded because of few person-time observations RESULTS: The incidence of ischemic stroke was 12.5, 8.7 and 10.8 per 1000 person-years for acenocumarol, dabigatran and rivaroxaban. After adjustment no differences were found for NOAC in the risk of ischemic stroke compared to acenocumarol (HR:0.78, 95%CI:0.51-1.20 for dabigatran; HR:0.76, 95%CI:0.42-1.35 for rivaroxaban). Regarding the incidence of mortality (80.2, 65.1 and 91.7, respectively) no increased risk was found (HR:0.96, 95%CI:0.81-1.12 for dabigatran; HR: 1.16, 95%CI:0.95-1.42 for rivaroxaban). The incidence of gastrointestinal bleeding was 18.3, 21.4 and 14.4, respectively, being dabigatran associated with a higher risk (HR:1.44, 95%CI:1.08-1.92) but not rivaroxaban (HR: 0.72, 95%CI: 0.43-1.19). However, no differences were found for major gastrointestinal bleeding (HR: 1.01, 95%CI: 0.59-1.73 for dabigatran; HR:0.43, 95%CI:0.16-1.17 for rivaroxaban). Regarding the incidence of intracranial haemorrhage (8.8, 3.0 and 5.4, respectively), dabigatran was found to be associated with a decreased risk (HR:0.36, 95%CI:0.18-0.74), and no differences for rivaroxaban (HR: 0.56, 95%CI 0.25-1.27). CONCLUSIONS: NOAC were not associated with an increased risk of stroke or death compared with acenocumarol among newly treated patients with AF. Dabigatran was associated with lower risk of intracranial haemorrhage and higher risk of gastrointestinal bleeding. The risk of major gastrointestinal bleeding related to NOAC was similar to that for acenocumarol.