Open defecation, which is still practiced by about a billion people worldwide, is one of the most examples of how place influences health in developing countries. Because of the negative healthy externalities of open defecation, eliminating it is a priority for governments and development agencies. Addressing the world’s remaining open defecation will require understanding why some people adopt latrines and others do not. In this study, we exploit the India Human Development Survey (IHDS), a nationally representative panel of households in India, where 60% of the world’s remaining open defecation occurs. IHDS households were surveyed in 2005 and again in 2012. Among rural households that defecated in the open in 2005, we investigate what baseline properties and what changes over time are associated with a household switching to latrine use between 2005 and 2012. We find that households that are richer or better educated, that have certain demographic properties, or that improved their homes over this period were more likely to switch to using a latrine or toilet. Unfortunately, however, each of these eff ect sizes is small; overall switching to latrine use from open defecation is low; and no ready household-level mechanisms are available for sanitation programs to widely influence these factors. Our research adds to a growing consensus that the social context should not be overlooked when trying to understand sanitation behavior. It also contributes to better research on the health consequences of poor sanitation: in order to conduct intervention studies, such as randomized experiments to learn about eff ects of open defecation on child height, researchers must first cause reductions in open defecation.