||Background: Information is needed on bleeding risk factors specific for patients with atrial fibrillation (AF) treated with non-vitamin K oral anticoagulants (NOACs). We aimed to identify risk factors in a large real-world cohort and to derive a bleeding risk score for patients with AF treated with NOACs. Methods: From nationwide registries (the Norwegian Patient Registry and the Norwegian Prescription Database), we identified patients with AF with a first prescription of a NOAC between January 2013 and June 2015. Cox proportional-hazards analysis was used to identify the strongest risk factors for major or clinically relevant non-major (CRNM) bleeding. Based on these, a risk prediction score was derived. Discrimination was assessed with Harrel's C-index. C-indexes for the modified Hypertension, Age, Stroke, Bleeding tendency/predisposition, Labile international normalised ratios, Elderly age, Drugs or alcohol excess (HAS-BLED), the Anticoagulation and Risk Factors in Atrial Fibrillation (ATRIA) and the Outcomes Registry for Better Informed Treatment of Atrial Fibrillation (ORBIT) scores were also calculated from the same cohort. Results: Among 21 248 NOAC-treated patients with a median follow-up time of 183 days, 1257 (5.9%) patients experienced a major or CRNM bleeding. Ten independent risk factors for bleeding were identified, which when included in a risk prediction model achieved a C-index of 0.68 (95% CI 0.66 to 0.70). A simplified score comprising three variables; age, history of bleeding and non-bleeding related hospitalisation within the last 12 months, yielded a c-index of 0.66 (95% CI 0.65 to 0.68). In the same cohort, the modified HAS-BLED, ATRIA and ORBIT scores achieved c-indexes of 0.62 (95% CI 0.60 to 0.63), 0.66 (95% CI 0.64 to 0.67) and 0.66 (95% CI 0.64 to 0.67), respectively. Conclusions: Our proposed simplified bleeding score could be a useful clinical tool for quick estimation of risk of bleeding in patients with AF treated with NOACs.