Abstract: Today the world’s fastest-growing cities lie in low-income countries, unlike the historical norm. Also unlike the “killer cities” of history, cities in low-income countries grow not just through in-migration but also through their own natural increase. First, we use novel historical data to document that many poor countries urbanized at the same time as the post-war urban mortality transition. Second, we develop a general equilibrium model of location choice with heterogeneity in demographics and congestion costs across locations to account for this. In the model, people prefer to live in low-mortality locations, and the aggregate rate of population growth and the locational choice of individuals interact. Third, we calibrate the model to data from a sample of poor countries, and find that informal urban areas (e.g. slums) grew in large part because they were able to absorb additional population relatively more easily than other locations. We show that between 1950 and 2005 the urban mortality transition by itself could have doubled the urbanization rate as well as the size of informal urban areas in this sample. Of these effects, one-third can be attributed to the direct amenity effect of lower urban mortality rates, while the remainder is due to higher population growth disproportionately pushing people into informal urban areas. Fourth, policy analysis suggests that family planning programs might be as effective as urban infrastructure and institution at transforming cities in poor countries into ``engines of growth’’