An increase of ^99mTc-sestamibi uptake in the myocardium during exercise was defined as a response rate, and the feasibility of a response rate for detecting coronary artery disease (CAD) was tested. Eighty-seven patients with suspected CAD had myocardial perfusion imaging with ^99mTc-sestamibi during exercise and at rest. A dose of 370 MBq of ^99mTc-sestamibi was injected at the maximal level of exercise, and a myocardial image was obtained 90 min later (exercise image). Then, 740 MBq of ^99mTc-sestamibi was administered at rest, and myocardial imaging was repeated (rest image). The exercise and rest images were corrected for physical decay and injected doses, and the exercise image was subtracted from the rest image to obtain the corrected rest image. A response rate was calculated as follows: (exercise image - corrected rest image)×100/corrected rest image (%). The global response rates of 20 patients without significant coronary stenosis (≤50%) were higher than those of 67 patients with significant coronary stenosis (81 ±33% and 50±28%, p<0.01). Global response rates were correlated with the maximal rate pressure products during exercise (r=0.56, p<0.01) and delta rate pressure products (r=0.53, p<0.01). Regional response rates in myocardial areas perfused by stenotic coronary arteries of ≤50%, 75%, 90% and 99-100% were 60±24%,* 56±33%,* 40±23%* and 30±23%,* respectively, (*p<0.01 vs without significant coronary stenosis). The response rates decreased as the severity of coronary artery stenosis advanced, and distinguished between coronary stenoses of graded severity. Accordingly, the response rate from myocardial perfusion imaging with ^99mTc-sestamibi may provide complementary information to the conventional inspection with myocardial tomography regarding the severity of CAD. (Jpn Circ J 1998; 62: 592 - 598)