This study was designed to assess how the precision requirements of discrete aiming movements affect the utility of brief visual samples provided during execution of movement. Subjects pointed with a hand-held stylus to targets with indices of difficulty of 3, 4, 5, and 6 bits with full vision, no vision, and in conditions in which 20-msec. visual samples were provided every 80, 140, or 200 msec. While intermittent vision required slightly longer movement times for targets with a high index of difficulty, subjects' accuracy was similar to the full-vision situation. Moreover, with intermittent vision, the movement trajectories resembled the full-vision and not the no-vision situation. It would appear that brief visual samples of the movement environment are sufficient for reasonably precise closed-loop control.