This paper sheds a new perspective on models of migration by integrating in a same structure the decisions about education and work; and by incorporating return migration with "brain waste". Brain waste is defined as the depreciation, due to migration, in the human capital acquired in the home country. This paper shows that brain waste, and return migration have significant effects on the optimal strategy of migration. Without these elements, brain drain is usually an optimal solution. But, when we introduce brain waste, then it is optimal to migrate as a student. Therefore, brain waste can explain why brain drain is not always optimal.