The antifibrillatory effects of flecainide 1 mg/kg + 0.05 mg/kg/min intravenously (i.v.), bretylium 6 mg/kg i.V., D-sotalol 2 mg/kg + 0.1 mg/kg/min i.v., and E-4031, a new class III drug, 50 mug/kg + 5 mug/kg/min i. v. were compared with three different methods of determining ventricular fibrillation threshold (VFT) in anesthetized open-chest dogs. In protocol 1, VFT was determined with 2-S, 50-Hz continuous pulses. Flecainide (n = 7) prolonged intraventricular conduction time (CT) and ventricular effective refractory period (ERP) and increased VFT significantly. Bretylium (n = 6) prolonged ERP slightly, but did not increase VFT significantly. Both D-sotalol (n = 6) and E-4031 (n = 6) prolonged ERP and increased VFT. In protocol 2, VFT was determined with the extrastimulus technique in dogs, with localized ventricular necrosis produced with protease. Flecainide (n = 10), D-sotalol (n = 8), and E-4031 (n = 8) restored VFT, which had been decreased by protease injection, to the baseline level, whereas bretylium (n = 8) did not. In protocol 3, the train pulse method with 100-Hz train pulses covering the vulnerable period was used in the same dogs used for protocol 2. Flecainide, bretylium, and D-sotalol increased VFT, but E-4031 did not. The antifibrillatory effects of class III drugs differ depending on the method of VFT determination. The present data suggest that the antifibrillatory effects of antiarrhythmic drugs should be assessed by different methods of VFT determination.