The perceived and remembered position of stationary target objects is subject to a large number of distortions. Objects are localized toward the fovea, and when an additional object (distractor) is presented, a tendency to average target and distractor position was observed. These distortions in visual short-term memory have been referred to as foveal bias and memory averaging, respectively. Because most studies on memory averaging did not monitor eye fixation, foveal bias and memory averaging may have been confounded. That is, observers may have fixated the distractor. To disentangle these factors, target and distractor were presented in the periphery, and fixation was monitored. Memory averaging was not observed. Rather a bias away from the distractor occurred when the distractor was briefly presented during the retention interval, or when it was visible throughout the trial. In contrast, a foveal bias was observed regardless of whether an additional object was present.