Nearly one sixth of the global population lives in urban “slums” – areas characterized by inadequate infrastructure and tenure security. This figure continues to grow as developing countries rapidly urbanize. Yet, the implications of these trends for urban poverty and social mobility are not well understood. While some argue slums provide temporary housing for rural migrants as they accumulate savings and eventually move to middle class neighborhoods, others argue slum residents are stuck in poverty traps. Deficits in longitudinal data on slums make it difficult to analyze the extent of social mobility. We iterate between satellite analysis and field knowledge to build an original sample of more than 9000 slum households across more than 200 slums from three Indian cities. To address the limitations inherent in cross-sectional data, we employ multiple methods and triangulate findings across household survey data, neighborhood focus group data, longitudinal satellite data, and in-depth qualitative interviews. While no one analysis is definitive on its own, all of these results point to the same conclusion: slum residents are neither stuck in poverty traps nor are they on a steady trajectory to joining the middle class. Movement out of neighborhoods, particularly to non-slum neighborhoods, is rare. Most households experience upward mobility within their neighborhoods, but the extent of improvement is capped at a low level, and, as opportunities increase, volatility increases in parallel. Plateauing and volatility are features present in low-end, and even more, in high-end slums. Engendering better livelihood opportunities requires reducing downward mobility while addressing the causes of plateauing upward mobility.