This dissertation presents a communication model of the impact of candidate character traits on vote preference and outlines how the communication environment surrounding US presidential elections influences the selection of salient candidate character traits, how these traits are framed by mediated communication, and reports their influence on vote preference which is dependent on voters’ level of news media use. A conceptualization of traits and the theoretical basis for how and why they may play a role in vote preference is presented and suggests traits provide an easily accessible heuristic allowing citizens to make predictions for future behavior of a candidate if elected. The theoretical foundation of the model is based on the integration of agenda-setting, priming and framing theories. The moderating role of media use on the relationship between perceptions of candidate traits and vote preference is also explicated. Support for this model is presented in three case studies that analyze three different datasets across two US elections. Specifically, the 2004 and 2008 general elections are explored using the 2004 and 2008 National Annenberg Election Studies (NAES) and an October 2008 dataset from Pew Research Center. These data provide empirical evidence supporting the theoretic argument driving this dissertation. All three case studies produce similar and consistent results and the congruence of the findings across different elections and multiple data sources contributes to the robustness and validity of the communication model of the impact of candidate character traits on vote preference offered in this dissertation.