Invasive electrophysiologic studies were performed in 102 patients with sustained ventricular tachycardia (VT) or ventricular fibrillation (VF) using an aggressive programmed electrical stimulation (PES) protocol. The study was repeated after 2.0 +/- 2.9 days in all patients with no intercurrent changes in antiarrhythmic therapy. Patients with coronary artery disease (n = 72) were identified and PES results of these patients were analyzed and compared with results of patients without coronary artery disease. Multiple clinical and electrophysiologic factors were analyzed to determine any association with concordance of PES responses. No significant difference in concordance of PES responses was found in the 2 groups of patients. PES responses were groups into 3 categories: (1) noninducible, (2) nonsustained VT, and (3) sustained VT. Kappa values of PES responses of noninducible and sustained VT in both groups were higher and therefore the PES responses were more reproducible than nonsustained VT. The induction of sustained monomorphic VT was more reproducible than a PES response of nonsustained or sustained polymorphic VT. Inducible sustained VT with a rate of greater than or equal to 250 beats/min was less reproducible than induction of sustained VT with a rate less than 250 beats/min. Induction of VT by 3 extrastimuli was less reproducible than with any other mode. This short-term variability may account for false negatives associated with PES-directed antiarrhythmic therapy. Because of these findings, it is recommended that nonsustained VT and sustained polymorphic or rapid polymorphic VT should not be used as PES end points to guide antiarrhythmic therapy.