We hypothesized that temporal lability in ventricular repolarization is a marker for, and is mechanistically related to, increased risk of malignant ventricular arrhythmias. To assess repolarization lability in the surface electrocardiogram, we developed an automated algorithm, based on template matching, to measure beat-to-beat changes in QT interval. We calculate a QT variability index (QTVI) to quantify the relative magnitude of QT interval changes compared to heart rate variability. We found that QTVI is a reproducible measure. It is elevated in patients with ischemic and nonischemic dilated cardiomyopathy compared with age-matched controls (P<.00001). We have also shown that QTVI is elevated in patients with malignant beta-myosin heavy-chain mutations associated with hypertrophic cardiomyopathy. In a study of patients undergoing electrophysiologic testing, QTVI identified patients with cardiac arrest better than electrophysiologic test result and better than other risk stratifiers included in the analysis. QT variability is a marker of electrical disease in the ventricle and may be associated with enhanced risk of life-threatening arrhythmias.