To compare the diagnostic accuracy between dobutamme echocardiography and treadmill exercise electrocardiography in detecting coronary artery disease in hypertensive patients, 43 patients without electrocardiographic evidence of left ventricular hypertrophy and basal ST-T changes, who had also undergone coronary angiography, were further evaluated by dobutamine echocardiography. The patients also underwent treadmill exercise echocardiography. Left ventricular mass index was calculated by echocardiography. Twenty-nine patients had coronary artery disease, of whom 22 had multi-vessel disease and 14 a normal coronary anatomy. Twenty-eight patients had an increased left ventricular mass index. The sensitivities of dobutaniine echocardiography and exercise electrocardiography for detecting coronary artery disease were 93% and 72% (P=0·08) respectively, and the specificities were 100% and 29%(P<0·005), respectively. Logistic regression analysis showed exercise electrocardiography to be a poor predictor of coronary artery disease (P<0·09) but dobutamine echocardiography was significantly better (P<0·00l). When patients with increased left ventricular mass index were excluded, prediction of coronary anatomy by exercise electrocardiography improved only marginally (P=0·4) while dobutamine echocardiography was significantly better (P<0·00l). Thus dobutamine echocardiography is superior to exercise electrocardiography for diagnosis of coronary artery disease in hypertensive patients.