Until recently, there were insufficiently long periods of rapid increase in incomes in a sufficient range of contemporary low income countries to study causes of poverty change. The exceptions were isolated micro studies. A decade or more of rapid growth in a wide range of countries has changed that situation. A recent compilation of data on poverty levels for a large number of countries allows study of factors associated with changes in poverty levels and facilitates conclusions about causality. Improvements in methodology have helped extract more information from the data than was previously possible. The initial basis for current cross section analysis was the continuing set of large scale sample surveys, carried out throughout India by the Indian Statistical Institute, over very diverse physical, economic, and social conditions, and finally extending into a recent period of rapid growth. This data set, using a similar methodology over time and across large areas provides a truly unusual opportunity for analysis of poverty relationships. Reinforcing the Indian data, is a large, cross national, inter temporal data set developed by Klaus Deininger and Lyn Squire (1996) that also includes ample number of countries with rapid economic growth and with a wide range of structures of that growth. In addition, data for a few Southeast Asian countries that have experienced rapid growth has become available, as has data for China covering its recent rapid growth. Together these data sets tell a detailed story of the sources of poverty and its decline in the context of growth. Initially, only the relation between overall growth and poverty was analyzed. The breakthrough in knowledge that shows it is the structure of growth that matters came from a series of empirical studies by Martin Ravallion and his colleagues at the World Bank. These studies examine the separate effect on poverty of growth by sector of the economy. Further studies, using a quite different methodology, by Peter Timmer, then at Harvard University, and his colleagues, provide independent corroboration of the basic relationships in the Ravallion studies as well as additional insights.