The majority of patients with serious ventricular arrhythmias induced by exercise have ischemic heart disease. These arrhythmias, however, develop only in a minority of the patients with coronary artery disease. The aim of this study was to investigate whether patients with ventricular tachycardia or fibrillation produced by exercise-induced ischemia exhibit any premonitory electrocardiographic indicators of arrhythmia propensity and whether arrhythmia suppression by myocardial revascularization abolished these changes. High-quality exercise electrocardiograms (50 mm/sec) from 30 case patients with ventricular tachycardia and fibrillation produced by exercise-induced ischemia were studied before and after surgical revascularization. These results were compared with those obtained from 30 control patients matched for age, sex, heart disease, and preoperative exercise capacity. The resting and peak exercise electrocardiograms were examined separately in a blinded manner with respect to QRS duration, ST-segment depression, and JT intervals. Patients with bundle branch block patterns were excluded. The QRS duration at rest was similar in case and control patients preoperatively and increased significantly with exercise in both groups. However, the QRS prolongation was larger in the case group, in which it was 11 +/- 3 msec compared with 4 +/- 2 msec in the control group (p = 0.043). QRS prolongation > or = 15 msec predicted ischemia-related ventricular arrhythmias in 73% of the patients. After surgical revascularization, there was no QRS prolongation with exercise in either group. In both groups, the QRS prolongation was associated with significant ST-segment depression, which was larger in the case patients.(ABSTRACT TRUNCATED AT 250 WORDS)