This chapter proposes a novel approach to measuring fund manager overconfidence and its impact on investment performance. Among numerous behavioural biases identified in financial agents, overconfidence is perhaps most widely studied. Abundant research suggests that overconfidence can have a significant value-diminishing impact on financial decisions taken by small investors, but very few studies have sought to measure the overconfidence of professional investors. This study investigates more than 4600 US equity mutual funds and demonstrates why the proxies commonly used to measure investor overconfidence cannot be readily applied to fund managers, hence the usefulness of the content analysis approach as an alternative. A number of potential proxies for overconfidence including overoptimism, excessive certainty, and excessive self-reference are measured across fund manager reports. The findings suggest that superior past performance boosts overconfidence, which is, in turn, associated with diminished future investment returns. In addition, word frequency analysis is conducted to shed light on the genre and tone fund managers employ in their investment narratives and how this interplays with their demonstrated level of overconfidence.