This article explores the spatial pattern of vulnerability to climate change hazards in Pakistan by developing a Human Vulnerability Index (HVI). For this purpose, we use Population Census 1998 and Agriculture Census 2000 data. The HVI places the 103 districts of Pakistan in rank order and looks at whether there is a correlation between human vulnerability and exposure to disaster of the districts with respect to climate change hazards such as floods. The HVI is further validated using an independent flood recovery data set. The study found that the HVI is a useful tool for identifying vulnerable regions and districts for resource allocation. But the HVI is a poor tool for vulnerability assessment at community and household levels. For this purpose we used logistic regression analysis, which indicates that the adult literacy rate, ownership of livestock, and access to electricity are the three (out of six) key variables that play a critical positive role in recovery after the 2010 floods. The primary data collected from households also reveal that the 2010 Pakistan floods have equally affecte d standing crops, livestock, and house structures. More than two-thirds of sample households had rebuilt their house structures, whereas livestock recovery was negligible since the floods. We also found that the 2010 floods affected some of the poverty regions of the country, but that there is a very weak systematic correlation between human vulnerability and disaster exposure.