In the 582 favelas (shantytowns) of Rio de Janeiro, the state has been historically absent; the police do not enter and even the most basic utilities, public transportation and schools are generally not provided to the inhabitants. In place of the state, gang organizations constitute the major political actor, having achieved a substantial degree of sovereignty through a monopoly on violence in these areas. Despite their similarities in most respects, gang organizations deal with residents under their control in vastly different ways. Some gangs implement responsive systems of law and justice, allow for democratic practices and maintain a high degree of social order while others implement more intrusive and unresponsive governing institutions while failing to provide social services or maintain social order. What accounts for these varying forms of governance implemented by gang organizations? Rio de Janeiro is the perfect laboratory to study gang governance as it remains the paradigmatic case of pervasive gang control in the urban sphere. Moreover, this project has important ramifications for urban public policy in much of the developing world in addition to engaging with important theoretical debates surrounding armed groups and their relationships to civilians. My dissertation will use two analytical approaches to investigate how and why gang organizations have engaged in varying forms of governance within these specific contexts. I will first implement a large-N quantitative analysis of the structural factors that exist in favelas that account for the variations in governance such as policing tactics, incursions by rival gangs, strength of civil society and the electoral value of the favela. This quantitative analysis will be followed by an in-depth qualitative study of four favelas, using case-studies, semi-structured interviews and resident surveys in order to provide another test of these hypotheses and effectively trace causal mechanisms.