Global climate change is likely to cause rising sea levels, more frequent and more powerful cyclones leading to greater storm surges along coasts, and increased frequency and severity of .flooding in low-lying areas (Solomon et al., 2007). While it is widely assumed by environmental scientists that these changes will be costly (Nicholls et al., 2007), there is little evidence on the magnitude or distribution of potential impacts. This project aims to provide evidence on this question, while also developing data sources that will be of general use. Bangladesh is believed to be one of the nations most vulnerable to climate change, in particular because of rising sea levels, increases severe storms and .flooding (Cruz et al., 2007). We seek to quantify the costs. To do so, we will create a comprehensive database on Bangladesh.s history of natural disaster events such as .oods, cyclones, and storm surges. We will then combine these data with data from the Census and household surveys to estimate the historical relationship between these natural disasters and socioeconomic outcomes. We will apply these estimates to projected changes in frequency and severity of natural disasters in order to predict the magnitude and distribution of impacts. In parallel, we will use current census and survey data to characterize the welfare e.ect of rising sea levels, in particular the loss in productivity associated with population displacement from the regions of Bangladesh projected to become permanently underwater. This project will directly inform climate change policies by providing detailed estimates of potential health and economic costs for the largest at-risk population in the world. Our results will help the Bangladeshi government and other interested parties better target mitigation policies across the various regions of the country. Climate Change and Flooding project will also have positive knowledge externalities, since we will provide new data on historical and projected .ood patterns, improve access to Bangladeshi health and socioeconomic outcomes data in one compiled source, and create a methodological template for similar data creation and analysis for other at-risk areas.