State interventions in local communities often have unintended consequences. While interventions are meant to improve residents’ lives—such as by reducing risk of exposure, helping in disaster recovery, managing the construction of the built environment—they may unintentionally cause additional harm. Local perspectives should be incorporated into planning and design because they often anticipate the likely impacts and unintended consequences of top-down interventions. Planners and policy makers have a number of tools to incorporate local perspectives, such as participatory planning, democratic representation, and social science methods. However, they are often under- or misrepresented, resulting in interventions with unintended negative effects. This dissertation contributes to two aspects of this challenge through research in Dar es Salaam and Moshi, Tanzania. First, efforts to incorporate local knowledge often fail to consider everyday practice of information use and flows between communities, government agencies, and civil society. Second, community risk perceptions differ from state- or expert-led risk assessments. The first challenge can be addressed by tracing information flows and stoppages between communities, government agencies, and civil societies. To address the second challenge, this dissertation presents methods to study two dimensions of community risk perceptions that differ from state-led approaches to understanding risk—social trust and place attachment. These chapters also offer methodological considerations for planners to consider that could better integrate local knowledge in planning processes and better represent local communities. The main findings suggest stoppages in the flows of local knowledge moving up the government hierarchy explain why local knowledge is often excluded or ignored in land use management decisions. To better integrate local knowledge in urban risk reduction efforts, planners need to attend to everyday governance practices and institutions. Regarding community risk perceptions, the research in Dar es Salaam shows that levels of social trust vary based on neighborhoods and between modes of representation, such as through government, civil society organizations, and interpersonal relationships. Representing local knowledge regarding social trust and place attachment helps to fill gaps in participatory and political modes of representation, especially when employed strategically considering the flows of information in practice.