The recent success of biventricular pacing with transvenously implantable left ventricular leads suggests that left ventricular leads may be useful for other modes of therapy. Animal studies showed small leads inserted into a left ventricular vein dramatically reduced defibrillation strength requirements. This article describes a human investigation of the feasibility of biventricular defibrillation. Fifty-one patients undergoing implantable cardioverter defibrillator (ICD) implantation were enrolled. After insertion of a standard ICD lead, a prototype over-the-wire left ventricular defibrillation lead was inserted through the coronary sinus and into a vein on the left ventricle. Lead insertion was guided by retrograde venography. The left ventricular lead's location was randomized to the anterior or posterior vein. Randomized, paired defibrillation threshold (DFT) testing was performed to compare a standard ICD shock configuration (Control: right ventricle- --> superior vena cava+ + CAN+) to 1 of 3 biventricular shock configurations. In the anterior vein, the left ventricular lead was tested with either a single biphasic shock from right ventricle + left ventricle- --> superior vena cava+ + CAN+ or a dual biphasic shock. In the posterior vein, the left ventricular lead was tested with a dual biphasic shock. Dual shocks consisted of a 40% tilt biphasic shock from right ventricle- --> superior vena cava+ + CAN+ followed by another 40% tilt biphasic shock from left ventricle- --> superior vena cava+ + CAN+, delivered from a single 225 microF capacitance. Left ventricular lead positioning was successful in 41 of 46 patients (89%). Mean left ventricular lead insertion time was 17 +/- 17 minutes and 13 +/- 15 minutes for anterior and posterior locations, respectively. Mean DFTs were not statistically lower for the left ventricular shock configurations, but retrospective analysis showed a well-defined region of the posterolateral left ventricle where consistent DFT reduction was achieved with dual shocks (14.0 +/- 2.7 J vs 7.8 +/- 0.9 J; n = 5; p = 0.04). There were no adverse events requiring intervention due to the use of the left ventricular lead. Biventricular defibrillation is feasible and safe under the conditions used in this study. Additional studies are needed to verify whether dual shocks with posterolateral left ventricular lead positions consistently reduce DFTs.