Foreign aid is likely the most important policy instrument employed by high-income countries to foster human development in low- and middle-income countries. In response to surging immigration pressure in Europe and the USA, policymakers have started promoting foreign aid to address the “root causes” of migration. However, in contrast to this apparent political consensus that foreign aid is effective, the scientific evidence on the effect of aid on development and migration is incomplete and inconclusive. We provide first global micro evidence on these effects by combining exceptionally rich survey data from a sample of almost one million individuals across the entire developing world with newly geocoded data on World Bank project allocation at the sub-national level over the period 2008-2019 and migration and asylum seeker flow data from the developing world to high-income countries. Using two independent causal estimation strategies, we show that, in the short-term (up to 2 years after the start of the project), foreign aid improves people’s aspirations in aid recipient areas and leads to a decrease in individual migration preferences and asylum seeker flows. In the medium-term (3 to 5 years after), foreign aid causes improvements in individual capabilities that translate into increasing regular migration flows to high-income countries. We provide extensive evidence on the heterogeneities of these effects by world region, aid target sector, and individual characteristics. Our results confirm that the “root causes” effect of aid exists, but is short-lived, and does not “work” in policymakers’ target areas, i.e., in fragile contexts. In contrast, foreign aid improves individual capabilities in the longer term which translates into increased regular migration as postulated by the “mobility transition”.