Honduras has exceptionally high rates of crime and violence, a great deal of which occurs in low-income, urban neighborhoods. Despite this, little scholarly attention has been given to how safe (or unsafe) the residents of such neighborhoods feel and how their levels of fear vary based on demographic attributes, perceptions of disorder, social relationships, and participation in local organizations. In this thesis, I leverage survey data collected in eleven low-income neighborhoods from across Honduras. Using a series of multinomial logistic models, I show that perceptions of social disorder and community cohesion and prior victimization are strong predictors of fear and that these variables provide greater explanatory power than the demographic attributes commonly referred to in studies from the United States and other developed countries. I then turn to the issue of collective efficacy, showing that individuals with higher levels of participation in community organizations are less likely to express feelings of fear, with this effect being the strongest for state-oriented - rather than civic or religious - organizations. This research serves to extend and deepen scholarly understanding of fear of crime in the unique context of Central America.