Introduction: Atrial fibrillation (AF) is more frequent in oncologic patients. Cancer and cancer therapies are risk factors for developing AF. The balance between thromboembolic and bleeding risks of AF in these patients is particularly challenging. Little is known whether embolic and bleeding risk scores used for general population can be applied in an oncologic setting. We aim to clarify if the scores are also valid for patients with breast cancer and AF. Methods: We included prospectively 465 women, 312 with AF & breast cancer and 153 with AF without cancer (mean age 73.86±14.16 year-old, mean follow-up 3.51±3 years). We analyzed parameters for the calculation of different embolic and bleeding risk scores (CHA2DSVASc2, HASBLED, ATRIA, SAMETT2R2, HEMORR2HAGES), and the antithrombotic therapy. We analyzed hemorrhagic and embolic events. Statistical analysis was performed with SPSS, and multivariate Cox regression analysis was performed to determine the independent predictive value of these scores in both groups.