This paper reports on a model of the sequential migration behaviour of some 76 000 Scottish and Welsh students, from their domicile location to the location of their higher education and on to their employment location. A logit model methodology is employed to analyse the choice of the location of the university attended, whether inside or outside Scotland or Wales. Then, within a GIS framework, migration-on-migration correlations and elasticities are estimated in order to identify the mobility effects of human capital acquisition. The results confirm the DaVanzo hypothesis that subsequent migration is related to previous migration and also the Sjaastad—Becker hypothesis that higher human capital individuals are more geographically mobile. However, there are institutional differences between the two countries which mean that the mobility effects of human capital acquisition have to be interpreted carefully in the light of other economic, geographical and social influences.