Catheter-directed laser injury of the left ventricular endocardium for ablation of ventricular tachycardia was studied in a canine model of simulated ventricular tachycardia. Bipolar plunge electrodes were placed at thoracotomy into the left ventricular endocardium in nine anesthetized dogs. Ventricular tachycardia was simulated by pacing at 200 beats per minute. After four days of recovery, catheter-directed neodymium:yttrium-aluminum-garnet laser injury was produced at the site of earliest recorded electrical activation during pacing tachycardia as detected by endocardial catheter mapping. Immediately after laser injury and again five days after injury, pacing tachycardia was attempted. Failure to pace after laser injury was defined as successful arrhythmia ablation. In 3/9 (33%) experiments, the laser-injured tissue surrounded the tachycardia source (pacing wires), and pacing-simulated ventricular tachycardia was prevented. When the laser injury did not involve the tachycardia source, 6/9 animals, due to limitations of the mapping system, pacing was not prevented. One animal developed sustained ventricular tachycardia during laser injury, with conversion to sinus rhythm by lidocaine. One animal, without recognized ventricular arrhythmia, died five days after laser injury. No unusual findings were noted at autopsy. These preliminary data suggest that catheter-directed laser-induced injury can ablate arrhythmia sources. Further studies are indicated in a more physiologic model, and the safety and risks of the procedure need further evaluation.