Lateralization in motor control, exemplified by handedness, is a publically observable lateral asymmetry. The understanding of causation of lateral asymmetry in movement is poor for two major reasons. First, structural limitations underlying lateral bias cannot be compelling in the sense that laterality is completely specified. Second, interacting sources of lateral bias are probably represented at several levels of the central nervous system, with some sources being many steps removed from the executing motor structures. Both factors permit flexibility in terms of which side does what. I suggest two approaches to improve understanding of mechanism. First, use of bimanual tasks which reduce flexibility in lateral expression and therefore provide more information than unimanual tasks. Second, analysis of task performances in terms of what levels of lateral specialization are likely to be involved. Such analyses will be especially useful in the understanding of hand use in left-handers who perform different skilled tasks with different hands.