Previous reports have suggested that dobutamine stress echocardiography compares favourably with other stress agent-imaging modality combinations for the detection of coronary artery disease. However, in daily clinical practice the value of a test is defined on a probability basis. To study the relative diagnostic contribution of clinical and dobutamine stress test variables, Bayesian analysis was performed in 223 patients with suspected coronary artery disease, who underwent coronary angiography and a high-dose dobutamine stress test in conjunction with electrocardiography, echocardiography and Technetium-99m sestamibi SPECT myocardial perfusion scintigraphy. According to the pre-test (clinical) probabilities, patients were divided into low-, intermediate- and high-risk groups; 155 patients were in the intermediate-risk group. After dobutamine stress echocardiography the number of patients in this intermediate-risk group was reduced to 102 (P < 0.0001). This reduction of patients in the intermediate-risk group by echocardiography was better than perfusion scintigraphy (102 vs 126 patients, P < 0.05) or classic markers of ischaemia such as angina and/or ST-segment changes (102 vs 150, P < 0.0001). Moreover, there was a good correlation between the echocardiographic post-test probabilities and the true distribution of coronary artery disease.