The concept of a 'fail-safe' device to ensure that sufficient current is delivered to render a pig immediately unconscious when electrically stunned was investigated. For live pigs, no significant correlation could be determined between the pre-stun low voltage sensed impedance and the actual, higher voltage stun impedance. In contrast, a good correlation was found using heads from pigs killed more than 24 h previously. The impedance of a live pig's head was predominantly a function of the stunning voltage and decreased non-linearly with increasing voltage. The 'difference' between live and dead pigs was attributed to an 'ageing effect' which is not significant in the first few hours post mortem. No change in phase angle between corresponding stunning current and voltage waveforms, indicated that tissue reactance was not a contributing factor. It is concluded that 'fail-safe' devices based on low voltage pre-stun sensing are unlikely to meet the current legislative requirements (Council Directive 93/119/EC).