The aim of the study was to examine the relation between the extent of myocardial ischemia and changes in QT interval dispersion in patients with obstructive coronary artery disease and in patients with normal coronary arteries. QT interval dispersion reflects regional variations in ventricular repolarization and cardiac electrical instability. Previous studies showed QT interval dispersion changes during episodes of myocardial ischemia in patients with coronary artery disease, but no data on the relation between extent of myocardial ischemia and degree of QT interval dispersion changes are available. To assess the effects of myocardial ischemia on myocardial repolarization by analyzing the change in QT dispersion during incremental atrial pacing, we studied 33 patients (7 women and 26 men, mean age 60.1 +/- 5.1 years, 18 patients with normal coronary arteries, 15 patients with coronary 3-vessel disease). QT dispersion was measured at baseline, after each pacing period, within 30 seconds after cessation of pacing ("peak ischemic stress"), and at 1-minute intervals for up to 5 minutes. Paired blood samples for determination of serum lactate were withdrawn from the coronary sinus and radial artery to determine the cardiac lactate extraction ratio at each point of electrocardiographic registration. In patients with coronary artery disease, QT dispersion increased from a baseline value of 39 +/- 7 ms to a peak ischemic stress value of 63 +/- 10 ms (p <0.0001). Patients with normal coronary arteries showed almost unchanged values of QT dispersion (41 +/- 9 vs 42 +/- 7 ms). There was a significant relation between the pacing-induced change in QT dispersion and the induced change in myocardial lactate extraction ratio (r = 0.76, p <0.0001). The change in QT dispersion (baseline vs peak pacing stress) was related to the extent of the cardiac lactate extraction ratio (r = -0.79, p <0.0001). These data indicate that the severity or extent of induced myocardial ischemia was related to the degree of induced changes of the variability in the timing of the ventricular recovery pattern.