Individual differences have been a part of leadership research since the days of trait theory. There has been an upsurge in interest in the topic due to recent theoretical and methodological advances. Also there is increasing interest inmeasuring individual leader differences using nontraditionalmethods such as the quantitative analysis of archival data.We describe a general methodology for developing and validatingmeasures of leader individual differences based on computerized language analysis of archival data. Two empirical examples focusing on narcissism among Fortune 100 CEOs illustrate the methodology. We summarize prospects and problems of computerized content analysis of archival materials.