The literature has traditionally focused on the impact of migration on the consumption of nonmigrants left behind, generally supporting the view that migration and remittances alleviate poverty. Another separate strand of research has yet provided evidence that those left behind bear a larger work burden to compensate for the loss of migrant’s earnings, thereby raising the concern of a negative effect on non-income dimensions of well-being, and in particular on leisure time. This paper aims at integrating these two approaches into one unified framework to account for the welfare impact of migration in both dimensions, i.e. in terms of both consumption and leisure time. Building on an agricultural household model, I derive a set of testable predictions necessary implied by migration decisions taken to maximize the family utility and driven by earnings’ differentials. Drawing on household panel data from rural Mexico and using fixed-effects instrumental variable estimation, I find evidence of both an income and a substitution effect which are both welfare-improving. Non-migrants both reduce their participation in off-farm jobs and increase their work in the household farm (and in self-employment in general). This reallocation of labor is consistent with a raise in the productivity of agricultural labor caused by the out-migration of a farmer (substitution effect). In line with a positive income effect, I find that nonmigrants do not completely offset the loss of migrant’s farm labor, which causes a decline in the total income derived from local activities at origin. Results further suggest that remittances sent by the migrant exceed his initial net contribution to the household income, thereby relaxing the budget constraint of non-migrants who increase their consumption.