Drawing on data from four African cities, this article argues that the primary determinants of urban protection have less to do with direct assistance and policy frameworks than individuals choices and positions in social and institutional networks. It demonstrates that legal status is neither consistently associated with particular migration histories nor a reliable predictor of effective protection. Furthermore, it finds that on aggregate, displaced people are not the most vulnerable urban residents. This has important implications regarding presumptions of vulnerability, the necessity of interventions aimed solely at refugees, and the political viability of refugee-centric programming. Lastly, it argues that effective strategies for promoting urban protection should be informed by the recognition that urban settlement is often, de facto, part of a long-term and even durable solution. Doing so means working to expand displaced peoples social, economic, and political choices. This may not be an attractive alternative to current practice for many humanitarian organizations, but working effectively in urban areas will require such a shift.