The effects of electrical stimulation of the vagus nerves and the administration of atropine on ventricular fibrillation threshold (VFT) were studied in open-chest hearts of 15 dogs anesthetized by alpha-chloralose. These studies were made in both normal and ischemic ventricles, i.e., before and during acute coronary occlusion. The ventricles were paces at a constant rate to eliminate rate-dependent changes and the minimal current required to induce ventricular fibrillation (or VFT) was determined by delivering a train of rapid rectanglular pulses (100 per second) to the venticle actoss the vulnerable period. In normal ventricles, VFT's were significantly increased by vagal stimulation (P less than 0.01) and decreased by atropine (P less than 0.05). Coronary occlusion markedly decreased VFT's (P less than 0.01), and vagal stimulation or atropine failed to alter VFT's significantly in these ischemic ventricles (P greater than 0.8). In additional 14 dogs, the effects of vagal stimulation and atropine were studied after the administration of propranolol. Propranolol alone increased VFT's significantly in boetreatment with propranolol, vagal stimulation and atropine failed to change VFT's significantly in both normal and ischemic ventricles (P greater than 0.8). These results indicate that the vagus nerves exert their effect on VFT by modifying the sympathetic nerve activity in normal ventricles, but such an effect is not significant enough to alter VFT in ischemic ventricles.