We examine the long-standing claim that understanding relational correspondence is a general component of representational understanding. Two experiments with 175 preschool children located in Norwich, United Kingdom, examined the use of a scale model comparing performances on a "copy" task, measuring abstract spatial arrangement ability, and the false belief task. Consistent with previous studies, younger children performed well in scale model trials when objects were unique (e.g., one cupboard) but poorly at distinguishing objects using spatial layout (one of three identical chairs). Performance was specifically associated with Copy task but not False Belief performance. Emphasizing the representational relation between the model and the room was ineffective. We find no evidence for understanding relational correspondence as a general component of representational understanding.