This dissertation is an economic study of household-level decisions related to flood risk mitigation. It is composed of four chapters that focus on the 2011 Thailand flood, the world’s most costly flood event in the past 30 years. The first chapter examines the magnitude and composition of economic costs that households in Bangkok bore during the 2011 flood. Two rounds of surveys with 469 Bangkok households collected detailed information on a broad set of flood costs. Results indicate that total flood cost was substantial. The median cost was equivalent to half of annual household spending. However, structural damage to homes was surprisingly low, given the depth and duration of the flood. The second chapter assesses how online information can enable households to reduce flood losses. Propensity score matching is used to test for evidence of a relationship between social media use and flood loss. Results indicate that social media use enabled households to reduce mean losses by 37%. Social media offered information that was not available from other sources, such as localized and nearly real-time updates of flood location and depth. With knowledge of current flood conditions, households could move belongings to higher ground before floodwaters arrived. The third chapter shifts focus to longer-term mitigation actions. It presents results from a randomized experiment that tests the effect of information on household uptake of flood insurance and home retrofits. A sample of 364 flood-prone households in Bangkok was randomly split into treatment and control groups. The treatment group received practical details on home retrofits and subsidized flood insurance as well as social norm information regarding insurance purchase decisions of peers. Results indicate that the information intervention increased insurance purchases by about four percent, while no effect was detected for home retrofits. The fourth chapter evaluates the social benefits of the information intervention presented in the third chapter. Results suggest that the intervention raises welfare of households, but not society. Furthermore, greater benefits are associated with better informing households that have high insurance demand, compared to using social pressure to persuade those with low demand.