In this article we study the effect of interregional migration on gross income of labour. The empirical analysis is based on Swedish household data for the period 1994 to 1995. We are primarily concerned with the allocation effect of migration on economic growth in an economy where the incentives for migration may be hampered by relatively high income taxes and transfers. A treatment-effect model for migration and income is estimated. The results indicate significant income gains from migration for the unemployed as well as for those who were employed prior to migration. The estimation results also indicate negative correlation between errors in the migration function and the income function. The hypothesis of no self-selection can be rejected.