Although microcredit is considered a crucial tool for economic development, its contribution to improving the welfare of poor households is not well established. This study investigates this issue. We start with a theoretical model in which microcredit is used to finance the family business of a representative household. The household maximises its lifetime utility, which yields a dynamic optimal consumption path as a function of microcredit borrowing. Guided by the theoretical model, we then empirically estimate the impact of microcredit using a panel dataset of rural Vietnamese households for the period 2008–2016. We find that microcredit programs in rural Vietnam were effective in raising household food consumption in both the short and long terms. Nevertheless, the burden of loan repayment may have diminished food consumption with a lag.