Study objective - The aims were to develop a Langendorff rabbit heart model and to compare monophasic and biphasic defibrillation pulses. Design - Hearts were perfused with a Krebs-Henseleit solution and two 1.4 cm2 Pt-Ir mesh patch electrodes were sutured onto the ventricles. A 5 ms monophasic or 10 ms biphasic pulse, with randomly selected voltages of 30, 50, 70, 90, 110, or 130 V, defibrillated the heart after 10 s of fibrillation. Subjects - 11 adult male New Zealand white rabbits weighing 2.8(0.27) kg, were used for the studies. Measurements and main results - A total of 72 fibrillation and defibrillation sequences were conducted in each preparation. The results were fitted to a sigmoidal dose-response curve by logistic regression analysis. Voltage and energy values from the fitted data at 50% and 80% success (V50, V80, E50, E80) indicated a significantly lower (p < 0.05) defibrillation threshold voltage and energy for the biphasic waveform [V50 = 48 (SD19) V, V80 = 87(27) V, E50 = 0.15(0.12) J, E80 = 0.48(0.29) J] compared with the monophasic waveform [V50 = 79(20) V, V80 = 110(20) V, E50 = 0.27(0.12) J, E80 = 0.5(0.12) J]. There was no observed difference in defibrillation success rate between the first and second halves of any study. Conclusions - The Langendorff rabbit heart model is suitable for assessing electrical fibrillation and defibrillation mechanisms. Defibrillation can be achieved with a lower energy when using a biphasic rather than a monophasic pulse.