The value of stress echocardiography in the routine diagnosis of coronary heart disease was assessed in 100 consecutive patients (22 women, 78 men; mean age 57 [32-83] years) scheduled for coronary angiography because of suspected angina. Exercise consisted of bicycle ergometry (n = 50), on the one hand, simulated exercise with transoesophageal atrial stimulation (n = 16), dipyridamole (n = 33) and dobutamine (n = 33) infusions, on the other. 91 patients were successfully tested by at least one of these procedures, while exercise electrocardiography was successfully employed in only 78 (P < 0.05). Stress echocardiography had a greater sensitivity than exercise electrocardiography (90% vs 78%) and specificity (90% vs 73%). Semiquantitative measurement of wall movement distinguished patients without functionally effective stenosis from those with single or multiple vessel disease (P < 0.001). Stress echocardiography thus makes it possible in most cases to demonstrate or exclude functionally significant coronary artery stenoses.