The value of serial electropharmacologic testing during long-term oral amiodarone therapy for prediction of long-term drug efficacy as well as characteristics of arrhythmia recurrence is controversial. One-hundred four consecutive patients with coronary artery disease and sustained ventricular tachyarrhythmias (VT) underwent initial electrophysiologic (EP) evaluation in the drug-free state and again after an amiodarone loading period of 25 +/- 14 days (mean +/- SD). Twenty-six patients (25%) had no inducible ventricular tachyarrhythmia during therapy with amiodarone (VT control group), whereas arrhythmia inducibility persisted in the remaining 78 patients (VT noncontrol group). During 17.4 +/- 13.7 months of follow-up, two patients in the VT control group either had VT recurrence or died suddenly compared with 21 VT recurrences and eight sudden cardiac deaths in the VT noncontrol group (actuarial event rates at 36 months of 0.11 and 0.56, respectively, p = .0065). The cycle lengths of recurrent VT in these 21 patients in the VT noncontrol group were compared with those observed at final EP testing. A significant linear correlation was demonstrated (r = .76, p = .0001). Subgroup analysis of patients in the VT noncontrol group showed no EP predictors of outcome, including cycle length of induced VT. However, patients dying suddenly during the follow-up period had a higher prevalence of new or worsening congestive heart failure (75%) compared with patients with VT recurrence (19%) or those with no arrhythmic event (29%) (p less than .02).2off