Objective: This study examined (1) the defibrillation efficacy of using a minimally invasive direct cardiac massage (MID-CM) device as one electrode of the defibrillation electrical circuit and (2) the effect on external defibrillation of defibrillating when the MID-CM device is in place and a pneumothorax is present. Methods: Part 1: in seven pigs, defibrillation thresholds (DFTs) were determined with a truncated exponential biphasic waveform. DFTs were determined for five electrode configurations: standard transthoracic defibrillation with electrodes on the left and right chest walls (1), with the MID-CM as one of the defibrillation electrodes pressed gently (2) or firmly (3) against the heart with the right chest wall patch as the second electrode, the same as (1) with the MID-CM device in place and the lungs at end-inspiration (4) or at end-expiration (5). Part 2: in six pigs, DFTs were determined with a monophasic damped sinusoidal waveform with external defibrillation electrodes (1) and with the device as one defibrillation electrode and the other electrode on either the anterior (2), lateral (3), or posterior right chest wall (4). Results: Part 1: the DFTs for (2) or (3) were not different (18.7±12.4 vs. 17.0±8.3 J), but both DFTs were lower than that for (1) (155±45 J). The DFT was elevated for (4) (205±69 J) compared with (1). For (5) only one animal could be defibrillated with shocks up to 360 J. Part 2: the DFTs for (2), (3) or (4) were not different (19.5±11.0, 25.4±9.4, 27.4±9.0 J), but all three were lower than the DFT for (1) (198±70 J). Conclusions: Using the MID-CM device as one electrode of the defibrillation circuit markedly lowers the DFT compared with that for standard transthoracic defibrillation for both a monophasic and biphasic waveform. Defibrillation with the device in place and the chest opened elevates the DFT for external defibrillation much more during end-expiration than during end-inspiration. © 2002 Published by Elsevier Science Ireland Ltd.