This study investigates circumstances that affect individuals' decisions of whether or not to flee their homes during civilian conflicts. Building on the “choice-centered” approach to studying forced migration, I test the argument that people make a decision to flee or stay even under highly dangerous circumstances. Using primary data collected through a public opinion survey in Nepal, I test a number of hypotheses regarding the impact of factors such as violence, economic opportunity, physical infrastructure or geographical terrain, and social networks on forced migration, providing an individual-level test of the choice-centered approach to studying forcedmigration. The empirical results are consistent with the major hypotheses developed in aggregate-level studies and provide better insights into the factors that affect individual-level behavior. Beyond conflict, there are a number of significant economic, social, physical, and political factors that affect individuals' choice to flee.