This paper takes a geographical approach towards assessing the ‘returns’ to family migration by addressing explicitly the impacts of differences in the cost of housing between the place of origin and place of destination for family migrants. While numerous studies have examined differences in labour-force participation and wages subsequent to migration, particularly on the part of wives, few studies have considered the local housing and labour market contexts of these events. This study examines the adjusted earnings outcomes resulting from migration for husbands, wives and families in the United States in the context of local housing costs. Our findings challenge the assumption of simple economic gains and instead indicate that who gains and who loses from family migration is quite complex. The geographical context of family migration is critical in determining gains and losses, and is interrelated with moves in and out of the labour market on the part of wives. Our research indicates that wives who leave the labour market after a move are very likely to have moved to a more affordable housing market. Conversely, wives are found to enter the labour market when the move is to a more expensive housing market. For this group, wives' earnings considerably mitigate the impact on overall family earnings. This paper provides an important contribution to understanding family migration by positioning the analysis of migration outcomes within the context of labour markets and local housing market costs. Copyright © 2006 John Wiley & Sons, Ltd.