The mechanism of action of moricizine, a new antiarrhythmic agent used in the Cardiac Arrhythmia Suppression Trial, is incompletely characterized. In addition, because moricizine is extensively metabolized, plasma moricizine concentration has an unknown relation to myocardial drug effect. Signal-averaged and standard electrocardiograms (ECGs) were used to monitor moricizine's myocardial effects in 16 patients with frequent ventricular premature complexes taking 600 to 900 mg daily. Three signal-averaged ECG variables were measured: total filtered QRS duration (fQRS), root-mean-square voltage in the terminal 40 ms of the QRS complex (V40) and the terminal low amplitude duration < 40-mu-V (LAS). At steady state, plasma samples were collected and serial recordings of signal-averaged and standard ECGs were taken at 0, 1, 2, 4, 6 and 8 h after moricizine administration. A 24 h ambulatory ECG was recorded throughout the test period. Moricizine prolonged the fQRS (p < 0.05) and decreased the V40 (p < 0.05) of the signal-averaged ECG and prolonged the QRS (p < 0.05) and corrected JT (JTc) intervals (p < 0.05) of the standard ECG. The time course of the signal-averaged and standard ECG variables paralleled plasma moricizine concentration; that is, the maximal changes occurred at 1 to 2 h and declined to time 0 values at 8 h. The maximal changes were: fQRS (+ 8%), V40 (- 33%), QRS (+ 8%) and JTc (+ 4%). Thus, dynamic changes were observed for intraventricular conduction (fQRS, QRS) and ventricular repolarization (JTc) over the dosing interval. In a subset of seven patients with greater-than-or-equal-to 6 ventricular premature complexes/h at steady state on moricizine, the frequency of ventricular premature complexes related inversely with plasma moricizine concentration and changes on the signal-averaged and standard ECGs. Noninvasive evaluation of moricizine provided insight into the relation between drug concentration and myocardial effect.