SUMMARY Fifty patients who presented with angina pectoris were studied to examine the disturbances of regional myocardial perfusion during stress. Each patient underwent 16-point precordial mapping of the ECG during an exercise test, and coronary and left ventricular angiography. Regional myocardial perfusion was assessed using an atrial pacing test and a short-lived radionuclide, krypton-81m. Eleven patients had negative exercise tests and uniform increases in myocardial activity of krypton-81m of 98 ± 18.0% during pacing. Ten patients performed 30,000-43,000 J in positive exercise tests. These patients showed abnormal coronary anatomy and increases in myocardial activity of krypton-81m to remote and jeopardized myocardium at the onset of pacing. However, further pacing produced a decrease in activity in the affected segment of 68.0 ± 9.0% accompanied by ST-segment depression and angina. Twelve patients achieved 26,000-32,000 J in positive exercise tests and had significant coronary artery disease. Atrial pacing produced increased activity of krypton-81m to remote myocardium. The jeopardized segment at first showed no change and then a decrease in regional activity of krypton-81m (89.0 ± 17%) accompanied by ST-segment depression and chest pain. Seventeen patients achieved only 7000-22,000 J in positive exercise tests. These patients showed abnormal coronary anatomy and developed decreases in regional activity of krypton-81m to the affected segment of myocardium starting at the onset of atrial pacing and decreasing by 88 ± 7.0% below control. We conclude that different patterns of disturbed myocardial distribution of krypton-81m are present during stress-induced ischemia in patients with coronary artery disease. There was a close temporal relationship between these disturbances and ST-segment depression. In this study we assessed disturbances of regional myocardial perfusion in patients with angina pectoris and coronary artery disease. Changes in the regional myocardial distribution of the short-lived krypton-81 m has been measured during atrial pacing and then related to the patient's symptoms, ECG during exercise and the findings at coronary arteriography.