Theories of attention have frequently pointed to the finding that there is a significant performance decrement ('cost') to responding to two different objects concurrently. However, much of the research aimed at investigating how attention is 'divided' in such circumstances has adopted response time (RT) as the measure of interest. In this paper we investigate how attention is 'divided' during the execution of concurrent motor responses, by studying bimanual reach-to-grasp movements directed towards two separate target objects. Furthermore, a key aspect of our study is that each hand is required to perform either the same action (congruent reaches) or a different action (incongruent reaches). Thus in Expt 1 we manipulated the movement amplitude of each hand, while in Expt 2 we manipulated object size. The results of this study suggest that while there is an overall cost associated with carrying out two movements simultaneously, kinematic measures are unaffected by whether the actions required of each hand are the same (congruent) or different (incongruent). The problem of executing incongruent bimanual movements appears to be solved by synchronizing each limb to a common movement duration, while movement velocity and grip aperture are independently scaled. These findings are discussed in relation to theories developed to explain the coordination of the reach-and-grasp phases of unimanual prehension, and in the context of recent theories of attention for action.