Antiarrhythmic effects of three new drugs, propafenone, tocainide, and SUN 1165, were examined using three canine ventricular arrhythmia models, i.e., digitalis, adrenaline and two-stage coronary ligation arrhythmias. The effects of procainamide, disopyramide, lidocaine, and phenytoin, class 1 antiarrhythmic drugs, on digitalis arrhythmia were also examined. The minimum effective plasma concentrations of all these drugs for each arrhythmia model were determined for a quantitative comparison. Propafenone and tocainide suppressed all the arrhythmias, while SUN 1165 suppressed digitalis and coronary ligation arrhythmias. The minimum effective plasma concentrations of propafenone for digitalis, adrenaline, 24-h coronary ligation, and 48-h coronary ligation arrhythmias were 1.8 +/- 0.7, 0.58 +/- 0.20, 3.5 +/- 0.3, and 3.6 +/- 0.9 micrograms/ml, respectively, and those of tocainide were 6.2 +/- 2.1, 23.7 +/- 9.0, 11.4 +/- 0.5, and 8.6 +/- 2.9 micrograms/ml, respectively (mean +/- standard deviation, n = 6-7). The minimum effective concentrations of SUN 1165 for digitalis, 24-h coronary ligation, and 48-h coronary ligation arrhythmias were 0.92 +/- 0.19, 2.5 +/- 0.4, and 1.2 +/- 0.4 micrograms/ml. The minimum effective concentrations for digitalis arrhythmias were 1.7 +/- 0.4 micrograms/ml for disopyramide, 10.1 +/- 2.4 micrograms/ml for procainamide, 3.5 +/- 1.6 micrograms/ml for lidocaine, and 11.3 +/- 3.0 micrograms/ml for phenytoin. Digitalis arrhythmia seems to be a useful model for detecting class 1 drugs, as it was suppressed by all the class 1 Na-channel blocking antiarrhythmic drugs, while class 2 beta adrenergic blockers and class 4 Ca-channel blockers had no effect. Also, not all the class 1 drugs suppressed coronary ligation and adrenaline arrhythmias.