The study examined the cognitive processes involved in switching from familiar to novel points of observation. Participants studied a single view of a small (table-top model) or large (laboratory) spatial display that was presented from a horizontal or vertical viewing angle. In a reaction-time task, spatial judgments were then made from imagined perspectives that were either the original perspective or a parallel, novel perspective. Familiar views resulted in decreased response latencies relative to novel views only for those participants who had been presented with a horizontal viewing angle, an effect that was independent of display size. The results indicate that viewing angle is a variable influencing people's ability to switch imagined perspectives.