The ACUTE trial randomly assigned patients who had atrial fibrillation (AF) of >2 days' duration to a transesophageal echocardiographically guided or a conventional strategy before cardioversion. In the 571 patients who underwent transesophageal echocardiography (TEE) in the ACUTE trial, we assessed the relative predictive value of baseline data derived by history, transthoracic echocardiography, and TEE for prediction of thrombus and adjudicated embolism (thromboembolism) as a composite end point. TEE was performed at 70 centers in 571 patients, 549 in the transesophageal echocardiographically guided group and 22 crossovers in the conventional group. Six patients (1.1%) who had embolism and 79 (13.8%) who had thrombi were identified in this group. Thrombus was completely resolved in 76.5% of patients who had repeat transesophageal echocardiographic procedures after 31.7 +/- 7.5 days of anticoagulation. For patients who had embolic events, none had a transesophageal echocardiographically identified thrombus; 5 of 6 (83.3%) had >/=1 transesophageal echocardiographic risk factors (including spontaneous echocardiographic contrast, aortic atheroma, patent foramen ovale, atrial septal aneurysm, mitral valve strands), and 4 of 6 (66.66%) had subtherapeutic anticoagulation or no anticoagulation. Clinical, transthoracic echocardiographic, and transesophageal echocardiographic risk factors contributed significantly to the prediction of composite thrombus/embolism. However, transesophageal echocardiographic thromboembolic risk factors were the strongest predictors of thromboembolism and provided statistically significant incremental value (chi-square 38.0, p <0.001) for identification of risk. Thus, in addition to thrombus identification, TEE has significant incremental value in the identification of patients who had high thromboembolic risk. In conclusion, this study supports the role of TEE and anticoagulation monitoring in patients who have atrial fibrillation and is useful for identifying thromboembolic risk factors.