Increasing flood losses in the Gulf of Mexico related to development patterns and climate hazards pose serious threats to resilience and insurability. The purpose of this study is to understand how scale, social vulnerability, risk, and urban form relate to National Flood Insurance Program (NFIP) policy coverage and flood exposure. Our multilevel models identify that flooding is significantly clustered by region and counties, especially shoreline counties. Our measures of risk suggest that the Federal Emergency Management Agency (FEMA) special flood hazard area (SFHA) underestimates risk and exposure when compared with the Flood Factor and that there is some compensation in terms of insurance coverage, suggesting a pattern of adverse selection. Older housing stock appears both less insured and less exposed, raising questions of whether current growth patterns are increasing risk independent of environmental change. Our models suggest that census tracts with higher percentages of black residents are less insured and more exposed, and a similar pattern exists for rural areas. Our results highlight the need to seek common solutions across the Gulf of Mexico, concentrating on the most flood-exposed counties, and that specific resilience strategies may be necessary to protect areas with socially vulnerable populations, especially in rural areas. Underlying challenges exist due to the spatial relationship between exposure and social vulnerability and the potential for adverse selection in insurance markets due to different measures of risk.