To evaluate the safety of uninterrupted rivaroxaban, a novel oral anticoagulant that directly inhibits factor Xa, and a vitamin K antagonist (VKA) in eligible adult patients with nonvalvular AF (NVAF) who are scheduled for a catheter ablation. This is a prospective, randomized, open-label, active-controlled, global multicenter safety study of up to 250 randomized patients. Eligible patients with paroxysmal or persistent NVAF, a left ventricular ejection fraction > 40 %, and a creatinine clearance > 50 mL/min will be randomized 1:1 to rivaroxaban 20 mg orally once daily or to dose-adjusted oral VKA (recommended international normalized ratio (INR) 2.0-3.0) and stabilized on anticoagulation therapy for 1-7 days (if no intracardiac thrombus on transesophageal echocardiogram (TEE) immediately prerandomization/post-randomization or if 3 weeks of sufficient anticoagulation is documented) or for 4-5 weeks (if no TEE, no documented 3 weeks of sufficient anticoagulation, or by patient choice). During catheter ablation, heparin will be administered (ACT-targeted range = 300-400 s) after catheter ablation, and VKA will be managed per usual care. The next dose of rivaroxaban will be provided at least 6 h after establishment of hemostasis. The primary endpoint will be the incidence of post-procedure major bleeding events observed during the first 30 +/- 5 days post-ablation. Secondary endpoints will include post-procedure thromboembolic events, additional bleeding, time-to-event, and medication adherence. This study is intended to provide information about the safety characteristics of rivaroxaban in patients with NVAF undergoing catheter ablation.