Abstract Exposures to extreme events are increasingly common in many parts of the globe as a function of changes in weather patterns combined with rising sea levels, but there is a paucity of data to support scientific research on the implications of such exposures for population health and well-being over the long-term. STAR, the Study of the Tsunami Aftermath and Recovery, is a unique exception. We have interviewed respondents from 10 months before the December 2004 Indian Ocean earthquake and tsunami for 10 years. This project will follow up the same respondents 15 years after the tsunami. The pre-tsunami baseline is a population-representative survey of 27,000 individuals who were living along the coast of Aceh and North Sumatra, Indonesia. The tsunami constitutes a large-scale, unanticipated natural disaster in an area that was not thought to be prone to tsunamis. Moreover, its impacts were spatially idiosyncratic across the study area. Whether a particular community was inundated by the waves depended on a combination of the wave direction, seabed and shoreline topography. Of 368 baseline communities, about a fifth were devastated, a third were somewhat damaged and the rest were not directly affected. These features of the natural experiment provide the basis for identifying causal impacts of this major natural disaster which killed 170,000 people in the study area. The baseline and six post-tsunami waves collected during the prior phase of this project provide detailed information about exposure to and experience of the disaster and the evolution of health and wellbeing, broadly defined, for baseline respondents who survived the tsunami plus new household members. In each follow-up, we have interviewed 95% of all baseline survivors. In this continuation we will re-interview all baseline survivors and their household members fifteen years after the tsunami. We will place these data in the public domain, adding to prior waves of STAR already in the public domain to create an unparalleled data resource for the scientific community. Focusing on those who were exposed to the tsunami as children (age <12y and in utero at the time), we will investigate the impacts of tsunami exposure on the evolution of health, education and cognitive performance, linking outcomes to changes in material and psycho-social resources at the family level as well as community level resources. The latter will be measured combining administrative data and information extracted from satellite imagery using machine learning. Longer-term impacts, fifteen years after the tsunami, will exploit innovative human capital measures including biomarkers, executive function and emotional reactivity. Loss of one or both parents is among the most extreme exposures. We will explore the causal effects of orphanhood by comparing children who lost parents in the tsunami with those in the same community who did not.