Existing studies suggest that leaders with previous rebellion participation have a higher level of international conflict propensity than leaders with no such experience. This dissertation examines whether prior rebel experience will induce leaders to initiate an international conflict in response to domestic strife. I propose a preference modification approach and argue that rebel leaders’ policy choice during domestic unrest is a product of their pre-existing preferences and contextual factors: contextual factors not only constrain leaders’ ability to pursue a certain policy but more importantly reshape their policy preference. Specifically, I claim that rebel leaders’ willingness to use force abroad during domestic unrest is contingent on the severity level of domestic problems. When rebel leaders face severe internal unrest, they are unwilling to engage in international conflicts because severe domestic strife will reshape leaders’ perceptions and neutralize their policy preference toward international conflict through two mechanisms. First, high intensity level of domestic strife changes the deliberative cost-benefit calculation about available policy options. Specifically, serious domestic problems call for a direct, speedy, and “to-the-point” policy response, which enables domestic measures (i.e. co-optation and repression) to be more efficacious because domestic measures aim to directly and effectively address the problem that gives rise to the strife. Second, severe domestic unrest affects rebel leaders’ intuitive behaviors by activating some certain predispositions of leaders endowed by rebellion experience, which induces them to use other policy responses rather than initiating an international dispute. Statistical analyses of the international militarized dispute initiation of leaders under domestic strife from 1875 to 2000 reveal strong support for these arguments. Two case studies of Mao Zedong and Suharto further confirm the causal mechanisms.