A bias in attention towards the dominant hand has been cited as a possible factor in the lateralisation of human bimanual coordination (Peters, 1981). A mirror was placed between the hands of 18 dextral participants performing rhythmic anti-phase movements. This set-up gave the appearance of a reflected virtual hand (moving in time with the un-occluded hand), in the same spatial location as the occluded left or right hand. This asymmetrical conflict between vision and action examined whether the left hand would show higher levels of error when replaced by a virtual right hand than the converse condition. Higher levels of error were observed during performance of the anti-phase pattern overall in the conditions where the mirror was present (compared to control conditions without the mirror). However, this effect did not differ between hands. The implications for the mirror paradigm, possible explanations for the lack of asymmetry, and the consequences for the attentional bias hypothesis are discussed.