The spatial and temporal organixation of unrestricted limb movements directed to small visual targets was examined in two separate experiments. Videotape records of the subjects' performance allowed us to analyze the trajectory of the limb movement through 3-dimensional space. Horizontal eye movements during reaching were measured by infrared corneal reflection. In both experiments, the trajectories of the different reaches approximated straight line paths and the velocity profile revealed an initial rapid acceleration followed by a prolonged period of deceleration. In Experiment 1, in which the target light was presented to the right or left of a central fixation point at either 10° or 20° eccentricity, the most consistent differences were observed between reaches directed across the body axis to targets presented in the contralateral visual field and reaches directed at ipsilateral targets. Ipsilateral reaches were initiated more quickly, were completed more rapidly, and were more accurate than contralateral reaches. While these findings suggest that hemispherically organized neural systems are involved in the programming of visually guided limb movements, it was not clear whether the inefficiency of the contralateral movements was due to reaching across the body axis or reaching into the visual hemifield contralateral to the hand being used. Therefore, in Experiment 2, the position of the fixation point was varied such that the effects of visual field and body axis could be disembedded. In this experiment, the kinematics of the reaching movement were shown to be independent of the point of visual fixation and varied only as a function of the laterality of the target position relative to the body axis. This finding suggests that the kinematics of a reaching movement are determined by differences in the processing of neural systems associated with motor output, after the target has been localized in space. The effect of target laterality on response latency and accuracy, however, could not be attributed to a single frame of reference, or to a simple additive effect of both. These findings illustrate the complex integration of visual spatial information which must take place in order to reach accurately to goal objects in extrapersonal space. Comparison of ocular and manual performance revealed a close relationship between movement latency for both motor systems. Thus, rightward-going eye movements to a given target were initiated more quickly when accompanied by reaches with the right hand than when they were accompanied by reaches with the left hand. The finding that the latency of eye movements in one direction was influenced by which hand was being used to reach suggests that reaching toward a target under visual control involves a common integration of both eye and hand movements.