R&D spillovers across firms have long been seen as a cause for faster innovation, with positive effects on productivity and growth (Griliches , Bernstein and Nadiri [1988, 1989], Romer ). Does inventors' mobility increase these spillovers? To answer this question we develop two alternative statistical tests. We exploit the time variation in patents' citations and transfers of inventors between Californian firms in the period 1980-2008. Our results provide strong evidence that, following the move of one inventor from onefirm to another, the R&D activities of the two firms become more interconnected.