Public Significance Statement We continuously make rapid eye movements called saccades to inspect the world around us. A short-term store is needed to hold this information temporarily, to build an internal representation of the visual scene, or guide our actions within the visual environment. It is well established that visual short-term memory is highly limited in capacity, which confronts us with a challenge: When storing information across saccade sequences, how do we reallocate memory resources to new incoming information? Here, we show that resources are dynamically reallocated away from previously fixated items stored in memory to make room for the new saccade target. A large amount of information is stored about this target, even before the saccade is made. Our results suggest a mechanism of how limited memory resources are dynamically allocated in order to support our ability to remember information from the visual scene. Little is known about how memory resources are allocated in natural vision across sequential eye movements and fixations, as people actively extract information from the visual environment. Here, we used gaze-contingent eye tracking to examine how such resources are dynamically reallocated from old to new information entering working memory. As participants looked sequentially at items, we interrupted the process at different times by extinguishing the display as a saccade was initiated. After a brief interval, participants were probed on one of the items that had been presented. Paradoxically, across all experiments, the final (unfixated) saccade target was recalled more precisely when more items had previously been fixated, that is, with longer rather than shorter saccade sequences. This result is difficult to explain on current models of working memory because recall error, even for the final item, is typically higher as memory load increases. The findings could however be accounted for by a model that describes how resources are dynamically reallocated on a moment-by-moment basis. During each saccade, the target is encoded by consuming a proportion of currently available resources from a limited working memory, as well as by reallocating resources away from previously encoded items. These findings reveal how working memory resources are shifted across memoranda in active vision.