The diagnostic accuracy of the dipyridamole test in provoking coronary insufficiency was investigated in 79 patients with chest pain and the results were compared with the findings on angiography and exercise electrocardiogram. 58 patients had documented severe coronary artery stenosis, 21 had patent coronary vessels (cardiomyopathy 8, aortic stenosis 1, ectopic origin of coronary artery 1, normal 11). Anginal pain after dipyridamole was a non-specific finding. Approximately half the subjects in whom coronary insufficiency would be expected according to the coronary angiographic and ventriculographic findings evidenced ischaemic ST-segment depression after dipyridamole, which was comparable to the number of positive exercise electrocardiograms. In 23 patients, most of whom had shown an inadequate frequency response during the initial exercise test, ergometry was repeated after the administration of dipyridamole. This resulted in an increase in ischaemic ECG response from 26 to 70%. It is concluded that a stress test combining dipyridamole and submaximum exercise increases the incidence of ischaemic ST-segment depression in comparison with ergometry alone. Anginal pain without ST-segment depression proved to be without diagnostic value.