Clinical role of $^{99m}Tc-MIBI$ myocardial scintigraphy in the diagnosis of coronary artery disease (CAD) is now well accepted, however, the role of it in the identification of viable myocardium in patients with chronic CAD has not yet been clarified. To determine the usefulness of rest-injected $^{99m}Tc-MIBI$ scan as a marker of myocardial viability, the regional uptake of this agent at rest was compared with that of $^{201}Tl$ on reinjection and 24 hours after reinjection images. Subject patients were 13 chronic CAD patients who showed irreversible perfusion defect(s) on standard pharmacologic (dipyridamole) stress-redistribution images. Immediately after the redistribution images were obtained, 37 MBq thallium was injected at rest, and images were reacquired at 10 minutes and 24 hours after reinjection. After then 740 MBq $^{99m}Tc-MIBI$ was injected, and 1 hour later rest MIBI myocardial imaging was performed. Five sets of imagestress, redistribution, reinjection, delayed images of thallium, and rest image of MIBI) were then analyzed qualitatively and quantitatively. Left ventricle was arbitrarily divided into 9 segments (apex, basal and apical portions of anterior, septal, inferior, and lateral walls). Seven patients and 30 regions showed a fixed perfusion defect on the stress-redistribution images. Among 30 regions, 15 showed positive uptakes and 6 showed negative uptakes on both $^{201}Tl$ reinjection/delayed images and $^{99m}Tc-MIBI$ rest images. Five regions showed only thallium uptake and were regarded as viable clinically. Of four regions which showed only $^{99m}Tc-MIBI$ uptake, two were regarded as viable, while the other two were regarded as a nonviable scar tissue clinically. In conclusion, $^{201}Tl$ reinjection technique was more reliable in the identification of viable myocardium. However, the role of $^{99m}Tc-MIBI$ in identification of viable myocardium was still remained to be clarified because 2 of 9 regions showed only $^{99m}Tc-MIBI$ uptake and were regarded as viable tissues.