We examined whether or not conscious knowledge about the availability of visual feedback on an upcoming trial would influence the programming of a precision grip. Twenty healthy volunteers were asked to reach out and grasp objects under two viewing conditions: full visual feedback (closed loop) or no visual feedback (open loop). The two viewing conditions were presented in blocked, randomized, and alternating trial orders. Before each block of trials, participants were explicitly informed of the nature of the upcoming order of viewing conditions. Even though participants continued to scale their grip to the size of the goal objects which varied in size and distance, they opened their hand significantly wider when visual feedback was not available during movement execution. This difference was evident before peak grip aperture was reached, continued into the grip aperture closing phase, and presumably reflects the visuomotor system's ability to build in a margin of error to compensate for the absence of visual feedback. The difference in grip aperture between closed- and open-loop trials increased as a function of distance, which suggests that the visuomotor system can make use of visual feedback given enough time, even when that feedback is not anticipated. The difference in grip aperture between closed- and open-loop trials was larger when the two visual feedback conditions were blocked than when they were either randomized or alternated. Importantly, performance did not differ between the randomized and the alternating trial blocks. In other words, despite knowledge of the availability of visual feedback on an upcoming trial in the predictable alternating block, participants behaved no differently than they did on randomized trials. Taken together, these results suggest that motor planning tends to optimize performance largely on the basis of what has happened regularly in the past and cannot take full advantage of conscious knowledge of what will happen on a future occasion.