As one of the most important additions to social science research on climate change in recent years, vulnerability studies have enriched our understanding of the factors that predispose people to climate-related risk. Yet, despite this enhanced understanding, engineering-based responses to hazards dominate contemporary policy and practice, while social causes are consistently overlooked. The proposed research uses Santa Fe, Argentina—a city ravaged by consecutive floods over the last decade—as a lens to explore why engineering-based responses persist despite their poor performance in reducing vulnerability. Due to a long legacy of hydraulic works, engineering-based responses to floods and flood risk have gone unquestioned by both government officials and residents and are deeply embedded in, rather than autonomous from, the political landscape. In contrast to the development literature that tends to pit the local in opposition to imposed state schemes, this dissertation project examines how state schemes work in conjunction with and through the practices and desires of local populations. By investigating how conventional hazards framing has come to dominate and shape practice in Santa Fe and how local political actors and residents are complicit in sanctioning engineering “fixes” that emerge from this framing, this research seeks to develop a textured, ethnographic understanding of how vulnerability is reproduced. In particular, it will expose how clientelism is both a pragmatic avenue for the urban poor to maintain their livelihoods as well as a key driver underlying why the poor continue to be vulnerable in the face of floods. Using insights derived from scholarship on urban geography, vulnerability and development, this project employs multiple methods, combining ethnography with survey, interview, and mapping techniques to investigate how vulnerability is reproduced through a tangled set of practices, processes, and relations interacting at multiple scales.