Multijoint movements such as reaching are impaired after brain lesions involving sensorimotor areas and pathways. However, the mechanisms by which such lesions affect motor control are not fully understood. Direct effects of the lesion may be partly compensated by both the system’s redundancy and its plasticity. Indeed stroke patients with limited arm movement can reach objects placed within the reach of the arm by using a compensatory strategy involving trunk recruitment. A similar strategy is observed in healthy individuals reaching for objects placed beyond the reach of the arm. Determining the control mechanism(s) governing this compensatory strategy in stroke patients was the goal of this study. Kinematics of reaching movements in hemiparetic and healthy participants to targets placed within and beyond the length of the arm were analysed. Targets were placed sagittally in front of the midline of the body. Two targets (targets 1 and 2) were within reaching distance defined as the length of the stretched arm from axilla to wrist crease. Two others were beyond arm’s reach so that one required a forward trunk inclination (target 3) and the other required body raising to a semi-standing position (target 4). Healthy participants used minimal trunk displacement for reaches to targets 1 and 2. For reaches to targets 3 and 4, trunk displacement increased with target distance. Whenever the trunk was involved, there was a stereotyped sequential recruitment of the arm and trunk in that the trunk began moving simultaneously with or before the hand and stopped moving after the end of hand movement. This suggested that the control system predicts that the trunk movement will be needed to extend the reach and includes the trunk, in an anticipatory way, into the reach. In contrast, most hemiparetic participants recruited their trunk for reaches to all four targets, even those placed close to the body. Similar to healthy individuals, the sequence of hand and trunk recruitment was stereotyped, suggesting that temporal planning aspects of the motor program underlying movement coordination were relatively unaffected. In contrast to healthy participants, the contribution of the trunk movement to the endpoint displacement was substantially higher in the hemiparetic group and occurred earlier in the reach. It is suggested that the target distance at which the trunk is integrated into the movement to extend the reach of the arm is attained around the limit of arm extension and that this limit is reduced in hemiparetic individuals.