When interacting with others, we often fail to anticipate factors in the situation or in our partners that may bias both how we interpret the intentions behind their behaviour and how they likewise analyze our actions and intentions. In particular, if we do not recognize how impoverished our behaviour is relative to our thoughts about our behaviour, we will greatly overestimate the ease with which our partners will be able to interpret this behaviour accurately. In this dissertation, two experimental demonstrations of people’s overestimation of their partner’s ability to read their behaviour are presented. In the first experiment, subjects were asked to finger-tap a popular tune of their choosing. Asked to estimate how likely it was that their listeners would be able to identify this tune, these subjects showed substantial overconfidence:​ tappers estimated that half the listeners would guess their tune;​ in reality, listeners were only able to identify two out of one hundred fifty tunes. Informed observers-people who knew what tune was being tapped but who had never served as tappers or listeners themselves-were also overconfident. They also estimated that fifty percent of listeners would be able to identify the tune. Male tappers and observers were more extreme than females in this failure to appreciate the listener’s perspective. This result is discussed in the context of broader gender differences in perspective-taking. The second experiment illustrated how people may similarly overestimate the ability of others to identify the intentions behind their behaviour in a social interaction. Results showed that people did, however, recognize how impoverished their perspective was in terms of figuring out their partner’s behaviour and were, therefore, less confident in their ability to read their partner’s intent Real-world implications of this bias are discussed.