The neural mechanisms underlying the visuomotor coordination of arm movements have been intensely investigated over the past 20-odd years (Georgopoulos, 1990; Kalaska and Crarnmond, 1992; Caminiti et al., 1996; Schwartz, 1994b). Most of the tasks used have involved movement of the arm toward stationary targets. Because the visual information about the target in such tasks is static, consisting simply of the target's location in space, the key corresponding movement parameters are the direction and amplitude of the movement. In real life, however, the visual target may change location, involving a dynamic aspect of visuomotor coordination.Previous studies addressed this issue partially by shifting the target location at various times during the reaction or movement time (Georgopoulos et al., 1981, 1983). With respect to monkey behavior, it was found that the hand moved first toward the first target for a period of time and then changed direction to move toward the second target. The duration of the movement toward the first target was a linear function of the time for which the first target remained visible. This result indicated that the arm motor system was strongly coupled to the visual system, and faithfully followed the changes in the location of the target. Similar results were also obtained in human subjects (Soechting and Lacquaniti, 1983). With respect to theneural mechanisms involved, it was found that cell activity in the motor cortex (Georgopoulos et al., 1983) and parietal cortex (Kalaska et al., 1981) changed promptly after the target changed location, from a pattern appropriate for a movement toward the first target to a pattern appropriate for a movement toward the second. Indeed, the duration of the first cell response was a linear function of the duration of the first target. Thus a strong and orderly influence of the visual condition was exerted on the neuronal activity in the motor and parietal cortex, and that influence was later reflected on the hand movement. The delay from shifting the target to the change of neuronal discharge patterns (from the first to the second pattern) was about 130 ms (see figure 10 in Georgopoulos et al., 1983). This value can be regarded as an estimate of the delay involved in the flow of information "on-line" from the visual to the arm motor system under behavioral conditions favoring a strong dependence of the latter on the former. In the monkey studies discussed above, the visual target was stationary and the monkeys were trained to move toward it. Under these conditions, the target shift served as a dynamic probe of visuomotor coordination.When the hand catches a moving target, however, we have a very different case of visuomotor coordination. Obviously, the drastic difference lies in the motion of the stimulus. In addtion, a richer set of task instructions is possible as well; for example, to catch the target as fast as possible or to catch it at a certain location (given a predetermined stimulus trajectory).Finally, a richer set of strategies by which the task can be accomplished becomes available; for example, when the target moves slowly and the task is to catch it at a certain location, the subject can wait and make a single catching movement, or can move incrementally toward the catching point by making a number of smaller movements. Experiments in both humans (Port et al., 1997; Lee at al., 1997) and monkeys (Port et al., 2001) showed that such different strategies can indeed be adopted. Results from singlecell recordings in behaving monkeys have been published (Port et al., 2001; Lee et al., 2001).