Several studies have questioned the efficacy of lidocaine in reducing the incidence of ventricular fibrillation shortly after acute myocardial infarction when arrhythmogenic mechanisms may be different from those operative several hours later. To determine whether lidocaine inhibits the occurrence of early ventricular fibrillation, the left anterior descending and septal coronary arteries were occluded at their origins in open chest anesthetized dogs. Fourteen of 16 control dogs died with ventricular fibrillation. Fifteen dogs received two different dose regimens of lidocaine before coronary occlusion. Of the 11 treated dogs maintaining lidocaine bl), 6 survived (P less than 0.05). Five dogs received the larger dose; all died, four having blood levels of 6.3 mug/ml or greater at the time of death. Ventricular fibrillation threshold also increased in six of eight dogs when lidocaine was administered after coronary occlusion. It is concluded that lidocaine at blood levels of 1.2 to 5.5 mug/ml significantly reduces the incidence of ventricular fibrillation early after coronary occlusion. Administration of this agent therefore may be of particular value in the early phase of acute myocardial infarction.