We measured the QRS duration during a treadmill exercise test and the QT interval using a 24-hour Holter electrocardiogram at various heart rates to identify use-dependent QRS prolongation and reverse use-dependent QT prolongation of class I and III antiarrhythmic drugs. Use-dependent QRS prolongation was detected in 61%, 53%, and 64% of patients receiving disopyramide, mexiletine, and pilsicainide, respectively. Reverse use-dependent QT prolongation was found in 40% and 70% of patients receiving disopyramide and E4031. Drugs suppressed > or = 75% of the total premature ventricular contractions in all patients who had both use-dependent QRS prolongation and reverse use-dependent QT prolongation, in 79% of patients with use-dependent QRS prolongation alone, in 70% with reverse use-dependent QT prolongation alone, and in 11% with neither use-dependent QRS prolongation nor reverse use-dependent QT prolongation. Use-dependent QRS prolongation and reverse use-dependent QT prolongation were identified noninvasively and were useful in evaluating antiarrhythmic efficacy.