The relationship between activity of the autonomic nervous system, myocardial ischemia, and malignant arrhythmias has been investigated in a new experimental preparation for sudden death. Fifty-seven dogs were chronically instrumented and studied under control conditions (n = 15) and 1 month after production of an anterior myocardial infarction (n = 42). The protocol consisted in occluding the left circumflex coronary artery for 2 min, commencing at the last minute of an exercise stress test and extending through the first minute after cessation of exercise. With this protocol, ventricular fibrillation was observed in 40% of normal dogs and 66% of dogs with infarction. In 14 dogs.with infarction, left stellectomy reduced the incidence of ventricular fibrillation to zero (p < .001). The reflex changes in heart rate elicited within the first minute of ischemia during exercise in the animals that survived (from 204 ± 14 to 198 ± 31 beats/min, -6) were opposite those in animals that had ventricular fibrillation (from 208 24 to 229 30 beats/min, + 21) (p < .05). The ischemiainduced reduction in heart rate despite continuation of exercise suggests the presence in the dogs that survived of active vagal reflexes that may have played an important role in the maintenance of cardiac electrical stability. This preparation has the potential to induce ventricular fibrillation consistently in conscious animals by the interaction of a few clinically relevant factors (acute myocardial ischemia, submaximal exercise and its cessation, sympathetic and vagal reflexes, and heart rate) and offers the possibility of acquiring further insights into the mechanisms of malignant arrhythmias and evaluating novel strategies for targeted prevention. Circulation 69, No. 4, 790-800, 1984. TARGETED prevention of sudden cardiac death depends on an adequate understanding of the mechanisms of and the factors leading to ventricular fibrillation, as well as an understanding of their relative roles. Basic electrophysiologic experiments, usually performed in isolated hearts or in open-chest preparations, are designed to identify these factors and their probable modes of action, but comprehension of their relative roles in the clinical setting requires experiments in unanesthetized animals.Necessary to achieve this goal are the identification of the most critical factors, the capability of controlling