The change in systolic time intervals from before exercise to three to four minutes following a maximal-exercise treadmill test was measured to eveluate chest pain in 110 fasting supine subjects. Forty-six (85 percent) of 54 patients with chest pain and with abnormal findings on coronary arteriograms were found to have at least a 10-msec prolongation in the left ventricular ejection time index (LVETI), whereas only two (8 percent) of 25 subjects without heart disease and 5 (16 percent) of 31 subjects with chest pain but with normal findings on coronary arteriograms had 10 msec or more of prolongation of the LVETI after exercise. The change in the other systolic time intervals (total electromechanical systole, preejection phase [PEP], and PEP/LVET) were less reliable in detecting the presence or absence of coronary disease. We conclude that determination of LVETI before and after maximal-exercise treadmill testing is a clinically useful noninvasive disgnostic test for obstructive coronary disease in patients with chest pain.