Available information on atherosclerosis of thoracic aorta in man is scanty and mostly derived from pathological or surgical series. Transesophageal echocardiography makes a clear definition of the entire thoracic aorta possible and enables large, population based studies. In order to define prevalence, risk factors and clinical implications of aortic atherosclerosis, the echocardiographic recordings of 220 patients suitable for both evaluation of thoracic aorta and risk factors analysis were reviewed. Transesophageal echocardiography has been performed because of valvular diseases (78), suspected aortic aneurysm or trauma (43), evaluation of valve prosthesis (39), previous cerebral or peripheral embolic events (22), infective endocarditis (14), cardiac mass lesions (12) or other indications (12). Age ranged from 5 to 81 years (55 +/- 15), male to female ratio was 0.99. Simple and complex atherosclerotic plaques were identified in 33% and 10% respectively. Complex atheromas were more frequent among patients with previous embolic episodes (6/22, 27% versus 17/198, 8.5%; p = 0.019). The prevalence of any type of atherosclerosis progressively increased from the fourth (8%) to the eighth (88%) decade of age. By univariate analysis age (p < 0.001), history of hypertension (p < 0.001), systolic (p < 0.001) and diastolic (p < 0.05) pressure, type II diabetes mellitus (p < 0.01), HDL cholesterol (p < 0.01), HDL/total cholesterol (p < 0.01) and uricaemia (p < 0.05) were associated with aortic atherosclerosis. Discriminant analysis identified 5 independent variables associated with the presence and the extent of atherosclerosis (Wilk's Lambda = 0.43): number of cigarettes per day, age, history of hypertension, systolic pressure and type II diabetes mellitus. This model provided a 63% correct classification rate.(ABSTRACT TRUNCATED AT 250 WORDS)