Background: There are two approaches to the treatment of atrial fibrillation: one is cardioversion and treatment with antiarrhythmic drugs to maintain sinus rhythm, and the other is the use of rate-controlling drugs, allowing atrial fibrillation to persist. In both approaches, the use of anticoagulant drugs is recommended. Methods: We conducted a randomized, multicenter comparison of these two treatment strategies in patients with atrial fibrillation and a high risk of stroke or death. The primary end point was overall mortality. Results: A total of 4060 patients (mean [±SD] age, 69.7±9.0 years) were enrolled in the study; 70.8 percent had a history of hypertension, and 38.2 percent had coronary artery disease. Of the 3311 patients with echocardiograms, the left atrium was enlarged in 64.7 percent and left ventricular function was depressed in 26.0 percent. There were 356 deaths among the patients assigned to rhythm-control therapy and 310 deaths among those assigned to rate-control therapy (mortality at five years, 23.8 percent and 21.3 percent, respectively; hazard ratio, 1.15 [95 percent confidence interval, 0.99 to 1.34]; P=0.08). More patients in the rhythm-control group than in the rate-control group were hospitalized, and there were more adverse drug effects in the rhythm-control group as well. In both groups, the majority of strokes occurred after warfarin had been stopped or when the international normalized ratio was subtherapeutic. Conclusions: Management of atrial fibrillation with the rhythm-control strategy offers no survival advantage over the rate-control strategy, and there are potential advantages, such as a lower risk of adverse drug effects, with the rate-control strategy. Anticoagulation should be continued in this group of high-risk patients.