How does inventors’ migration affect international talent allocation, knowledge diffusion, and productivity growth? To answer this question, I build a novel two-country innovation-led endogenous growth model, where heterogeneous inventors produce innovations, learn from others, and make dynamic migration and return decisions. Migrants interact with individuals at origin and destination, diffusing knowledge within and across countries. To quantify this framework, I construct a micro-level dataset of migrant inventors on the US-EU corridor from patent data and document that (i) gross migration is asymmetric, with brain drain (net emigration) from the EU to the US; (ii) migrants increase their patenting by 33% per year after migration; (iii) migrants continue working with inventors at origin after moving, although less frequently; (iv) migrants’ productivity gains spill over to their collaborators at origin, who increase patenting by 16% per year when a co-inventor emigrates. I calibrate the model to match the empirical results and study the impact of innovation and migration policy. A tax cut for foreigners and return migrants in the EU that eliminates the brain drain increases EU innovation but lowers US innovation and knowledge spillovers. The former effect dominates in the first 25 years, increasing EU productivity growth by 3%, but the latter dominates in the long-run, lowering growth by 3%. On the migration policy side, doubling the size of the US H1B visa program increases US and EU growth by 4% in the long-run, because it sorts inventors to where they produce more innovations and knowledge spillovers.