The purpose of this study was to determine if body stimuli are uniquely processed by the visual recognition system. First, my results supported past findings showing that body processing differs from object processing (e.g., cars, chairs, houses). However, body processing depended on the presence/absence of a head. Second, the nature of appearance impacted observers’ performance such that discrimination was better for real than artificial bodies. Finally, I examined the impact of body appearance on event-related potential (ERP) responses, specifically the P100 and N170, and found that amplitudes elicited by real headless bodies was significantly larger than amplitudes elicited by all other variations in body appearance. In general, these results suggest the existence of a body recognition system that processes body images varying in visual appearance. However, this system may be more tuned to bodies that most resemble natural appearance and less tuned to bodies that deviate away from it.