Objective. To observe the safety and efficacy of warfarin and rivaroxaban in anticoagulation therapy in patients with atrial fibrillation (AF). Methods. A total of 96 patients with AF treated in our hospital from June 2019 to February 2021 were enrolled in this study. According to the different modes of drug administration, the patients were divided into the warfarin group and rivaroxaban group. Demographic and clinical data such as age, body weight, and previous drug use were collected. The blood routine, liver and kidney function, blood coagulation routine, and cardiac color ultrasound were accessed. The valvular atrial fibrillation and anticoagulant taboos were excluded, and the risk of embolism and bleeding was evaluated. Among them, 48 patients in the warfarin group were given warfarin once a day, and the international ratio (INR) was used to adjust the dose, and the INR was controlled between 2.0 and 3.0. In contrast, 48 patients in the rivaroxaban group received a fixed dose of rivaroxaban 20 mg or 15 mg once a day. After administration, regular telephone or outpatient follow-up was given once a month, to monitor patients’ drug compliance and ask if there was bleeding, and to detect blood routine, urine routine, fecal routine+occult blood, and liver and kidney function. In addition, at the beginning of 3, 6, and 12 months of follow-up, each patient was given cardiac color Doppler ultrasound, peripheral vascular color ultrasound, and brain CT to determine whether there were mural thrombosis, stroke, and peripheral arterial thromboembolism. The INR attainment rate, coagulation index, thromboembolism, bleeding, and adverse reactions were compared between the two groups. Results. There was no significant difference in serum Dmurd and NT-proBNP levels between the two groups before treatment and 3, 6, and 9 months after treatment. There was no significant difference in the number of venous embolism, pulmonary embolism, cerebral embolism, and total embolism between the two groups ( P > 0.05 ). There was no significant difference in the number of mild, moderate, and severe bleeding between the two groups ( P > 0.05 ), but the total number of bleeding in the rivaroxaban group was lower than that in the warfarin group ( P < 0.05 ). During the treatment, side effects such as nausea and vomiting, elevated transaminase, glutamyl transpeptidase, and diarrhea occurred between the two groups, and there was no significant difference in the number of adverse reactions between the two groups ( P > 0.05 ). Conclusion. Compared with warfarin, rivaroxaban anticoagulant therapy has the same advantage in tolerance and prevention of thromboembolism in patients with AF, but rivaroxaban can effectively reduce the risk of bleeding in patients with AF.