In the normal population the duration of electrical systole (QT) is shorter (mean -26 +/- 13 ms, 1 standard deviation) than that of electromechanical systole (QS2), which it closely parallels throughout the range of resting heart rate. This close association prompted the hypothesis that the duration of QT may be more closely linked physiologically to QS2 than to the heart rate and hence may provide a more potent prognostic indicator than the relation of QT to heart rate. The present study was designed to test this hypothesis. On hundred stable patients 14 months after myocardial infarction were followed up an average of 43 months. Twenty patients had prolongation of QT relative to QS2 (QT greater than QS2) and 13 patients had long QT corrected for heart rate (QTc). There were 20 deaths (16 sudden). The cumulative 5 year survival rate was 35% in patients with QT greater than QS2 and 91% in patients with QT greater than or equal to QS2 (p less than 0.001). There was no significant difference in survival in patients with long QTc compared with patients with normal QTc (69 versus 82%). Univariate and multivariate analysis for the association of QT greater than QS2 with known risk factors, clinical descriptors, drug therapy, systolic time intervals (preejection period/left ventricular ejection time), and frequency of 1, 2 and 3 vessel coronary occlusive artery disease revealed that only the presence of abnormal preejection period/left ventricular ejection time added significantly to the presence of QT greater than QS2 in stratifying 5-year survival into high and low risk groups. Thus, in patients with coronary artery disease, the presence of QT greater than QS2 provides a new risk indicator that is more potent than QT corrected for heart rate.