Amputation of a limb induces changes in the so-called body schema, which might be influenced by the use of prosthetic devices. Changes in the body representation associated with prosthesis use could be investigated using a hand mental rotation task. However, direct neurophysiologic evidence for the effect of prosthesis use on hand mental rotation is still lacking. In this paper, we recruited two groups of unilateral upper-limb amputees, i.e., amputees using a prosthesis or with a history of prosthesis use (Pro group) and amputees without a prosthesis (non-Pro group), as well as a sample of matched healthy controls. Using concurrent behavioral and electrophysiological assessments, we found that Pro amputees were comparable to healthy controls in either behavior or event-related potentials (ERPs), while non-Pro amputees showed prolonged response time as well as divergent ERP patterns. The P200 amplitude of non-Pro amputees was significantly larger for the non-dominant hand pictures than that for the dominant hand pictures, while such a hand difference in P200 was not found in either healthy controls or Pro amputees. Furthermore, the typical angular modulation of the N200 amplitude in healthy controls and Pro amputees was not presented in non-Pro amputees. Our results suggest that prosthesis use could preserve mental rotation ability by maintaining the performance of motor imagery and visual perception of hands, which represents a preservation of the body schema.