Identification of dipyridamole-induced regional wall motion abnormalities by echocardiography has recently been proposed as an alternative diagnostic stress test for coronary artery disease. This study evaluates this new technique by comparing the results obtained (overall, regionally and by abnormality type) with those of thallium-201 myocardial imaging after dipyridamole stress in 25 patients. Acceptable echocardiograms were obtained in 20 patients (80%). Concordance of echocardiographic abnormalities for both overall and regional thallium abnormalities was 85%. Sensitivity, specificity and predictive value of dipyridamole echocardiography for overall and regional thallium defects were 92%, 71% and 85%, and 91%, 81% and 85% respectively. However, concordance between the two for abnormality type (i.e. ischaemia versus infarction) was only 66% and the sensitivity, specificity and predictive value of dipyridamole echocardiography for identifying ischaemia as opposed to infarction were only 43%, 82% and 63%, respectively. There was substantial agreement between thallium and echocardiographic imaging after dipyridamole infusion in the diagnosis of coronary artery disease. Echocardiography appears less well able to distinguish infarction from active ischaemia. Dipyridamole echocardiography provides a highly versatile, noninvasive bedside stress test for the detection and localization of coronary artery disease.