Theory of mind-or the understanding that others have mental states that can differ from one's own and reality-is currently measured across the lifespan by a wide array of tasks. These tasks vary across dimensions including modality, complexity, affective content, and whether responses are explicit or implicit. As a result, theoretical and meta-analytic work has begun to question whether such varied approaches to theory of mind should be categorized as capturing a single construct. To directly address the coherence of theory of mind, and to determine whether that coherence changes across development, we administered a diverse set of theory of mind measures to three different samples: preschoolers, schoolaged children, and adults. All tasks showed wide variability in performance, indicating that children and adults often have inconsistent and partial mastery of theory of mind concepts. Further, for all ages studied, the selected theory of mind tasks showed minimal correlations with each other. That is, having high levels of theory of mind on one task did not predict performance on another task designed to measure the same underlying ability. In addition to showing the importance of more carefully designing and selecting theory of mind measures, these findings also suggest that understanding others' internal states may be a multidimensional process that interacts with other abilities, a process which may not occur in a single conceptual framework. Future research should systematically investigate task coherence via large-scale and longitudinal efforts to determine how we come to understand the minds of others.3