Twenty-six patients (19 men and 7 women) with symptomatic ventricular tachycardia (VT) were studied using invasive and noninvasive techniques to induce VT. Of the study population, 12% had syncope and VT on Holter monitoring, 30% had cardiac arrest and 58% had symptomatic VT. All patients had antiarrhythmic agents stopped 5 half-lives before evaluation and then had autonomic profile (upright tilt, cold pressor test, exercise testing and hand grip) as well as programmed electrical stimulation studies performed. Autonomic profile testing induced VT in 5 of 26 patients (19%) and in only 1 patient was the arrhythmia reproducibly induced. All 26 patients had VT induced on electrophysiologic testing; 9 patients had nonsustained and 17 had sustained VT. Lorcainide administered intravenously prevented VT induction in 20 of 26 patients tested, whereas procainamide was effective in 11 of 24 patients. Ten of the 13 not protected by procainamide were protected by lorcainide. Twenty patients were started on long-term lorcainide therapy and followed up for 29 +/- 3.4 months. Five patients have discontinued therapy, 2 because of breakthrough arrhythmias, 2 because of severe sleep-wake disturbances and 1 because of private physician preference. An additional 3 patients died during therapy because of myocardial infarction in 1, progressive myopathy in 1 and sudden death in 1. Sixty percent of patients started on lorcainide therapy have continued. In this patient population, noninvasive induction of VT is not a sensitive or reproducible technique in assessing antiarrhythmic therapy. Furthermore, when selected on the basis of electrophysiologic testing, lorcainide is a well-tolerated and effective antiarrhythmic agent.