Background <p>Rivaroxaban and edoxaban are commonly used for stroke prevention in patients with atrial fibrillation (AF). This study aimed to compare the bleeding risk between rivaroxaban and edoxaban in patients with atrial fibrillation, both with and without the combination of amiodarone.</p> Methods <p>In this prospective cohort study, participants were divided into four groups: rivaroxaban monotherapy, edoxaban monotherapy, rivaroxaban combined with amiodarone, and edoxaban combined with amiodarone. Propensity score matching (PSM) was used to control for confounding factors. The primary endpoint was the time to the first occurrence of bleeding (major, CRNM, or minor).</p> Results <p>A total of 910 patients were included. The bleeding risk was similar between the rivaroxaban (9.0%) and edoxaban (13.5%) monotherapy groups (<i>P</i> = 0.403). However, the bleeding risk in the rivaroxaban-amiodarone group was significantly higher than in the edoxaban-amiodarone group in the PSM analysis (HR = 2.015, 95% CI = 1.026–3.958, <i>P</i> = 0.042). Patients receiving standard-dose rivaroxaban with amiodarone had a significantly higher bleeding risk than those receiving edoxaban with amiodarone (HR = 2.678, 95% CI = 1.017–7.045, <i>P</i> = 0.046). Additionally, rivaroxaban with amiodarone was associated with a significantly higher bleeding risk compared to edoxaban with amiodarone in patients with renal dysfunction (HR = 2.186, 95% CI = 1.016–4.702, <i>P</i> = 0.045).</p> Conclusion <p>When combined with amiodarone, rivaroxaban was associated with a higher bleeding risk compared to edoxaban, particularly in patients receiving the standard dose of anticoagulants or with impaired renal function.</p> <p><b>This study the name of the trial register</b>: Zhongshan Hospital, Fudan University, <b>Registration number</b>: ChiCTR2200062662, Date of Registration: 2022-08-15.</p>