Should individuals migrate before acquiring education or after? In order to analyze the optimality of the timing of migration, I develop a model of migration, which combines the two migration decisions into a unique model - decisions about where to get an education and decision about where to work. The main reason for having a unified model is that investment in human capital cannot be disjoined from the decision about work. This paper shows that brain drain is usually an optimal solution. But, when we incorporate "brain waste" and "return migration", then it is optimal to migrate when young.