The diagnostic accuracy of transesophageal echocardiography and its relevance on the decision making process were evaluated in 80 patients with suspected aortic dissection during a 37-month period. The diagnosis was proven by aortography and/or magnetic resonance and/or computerized tomography and/or surgery and/or autopsy in each case. Transesophageal echocardiography had no serious complication. A correct diagnosis of aortic dissection was made in 39 of 40 patients (sensitivity, 97.5%; specificity, 100%) and the type of dissection was correctly demonstrated in each case. Thrombi in the false lumen were detected in 16 patients. The primary entry site was correctly identified in 33 patients (85%). Aortic regurgitation was detected in 25 cases (severe in 9) and pericardial effusion in 14 (with tamponade in 2). Transesophageal echocardiography was more accurate than aortography in the diagnosis of noncommunicating intramural dissection (2 cases) and identified more precisely the retrograde extension of the dissection in DeBakey type III patients (4 cases). In 9 cases surgical indication was based on clinical data and transesophageal echocardiography alone. We conclude that transesophageal echocardiography allows a bedside, safe and accurate diagnosis of aortic dissection. In the majority of the patients it provides the minimal diagnostic information necessary to the therapeutical decision making.