In a multicenter randomized trial, we studied a transesophageal echocardiography (TEE) guided strategy with short-term anticoagulation compared with a conventional strategy for patients with atrial fibrillation >2 days' duration and undergoing cardioversion. Composite major and minor bleeding was a predetermined secondary end point of the study. The objective of the study was to assess the incidence, location, and predictors of bleeding in the 2 treatment groups. A total of 1,222 patients were assigned to a TEE guided or conventional strategy and followed over 8 weeks. We present data on major and minor adjudicated bleeding complications for the 2 study groups during the 8-week study period. Composite major and minor bleeding complications occurred in 51 of 1,222 patients (4.2%) and were significantly lower in the TEE guided group compared with the conventional group (2.9 vs 5.5%, p = 0.025). The TEE group had fewer cancellations of cardioversion as a result of bleeding (0% vs 0.7%, p = 0.003). Major (n = 14) and minor (n = 38) bleeding complications were predominantly gastrointestinal (71.4% and 31.6%, respectively) and were associated with warfarin use. Predictors of bleeding included patient age, conventional group assignment, inpatient status, and functional status. Thus, composite major and minor bleeding complications occurred in 4.2% of the 1,222 patients and were significantly lower in the TEE guided group compared with the conventional group. Treatment variables affecting length of anticoagulant therapy in the conventional arm combined with advancing age and functional status are important concerns in patients who undergo cardioversion of atrial fibrillation.