An implication was examined of three assumptions of Personal Construct Theory. Personal and supplied construct systems were compared for their capacity to account for 34 subjects' behaviour in an independent sorting task. It was found that either construct system tended to account for a significant amount of the variance in sorting behaviour, but that more was accounted for by personal Construct systems. Observations were made on the way in which this phenomenon varied between subjects, sub-tasks, and parts of construct systems. The assumptions examined were held to be relevant to Social perception and cognitive theories.