The existing literature on forced migration limits our understanding of how violence affects migration to competing destinations. This article adds to the literature on forced migration by studying how armed violence during a period of civil conflict in south-central Nepal influenced the likelihood of local, internal, and international migration. We find that violence has a nonlinear effect on migration, such that low to moderate levels of violence reduce the odds of movement, but when violence reaches high levels, the odds of movement increase. We also find that the effect of violence on mobility increases as the distance of the move increases. When we consider the influence of violence on microlevel decision-making, we find that the effects of individual and household-level determinants were mostly consistent with hypotheses derived from contemporary theories of voluntary migration and that no predictor of migration influenced the decision to migrate differently in the presence of violence.