We develop a dynamic spatial equilibrium model to capture the role of wealth in moving decisions. The model’s features are motivated by novel evidence on heterogeneous mobility patterns by ability to borrow. Less able to borrow tend to move more but are less likely to move to “opportunity". We show that by introducing consumption-saving decisions, the model is able to generate heterogeneous migration patterns consistent with the data without imposing different migration costs across demographic groups. We perform counterfactual analysis: i) we compare sources of mobility restrictions and find that utility vs monetary moving costs shape the migration-wealth relationship in opposite ways; ii) a moving subsidy to unemployed in the Canadian “Rust Belt" increases welfare by 0.4% and most of it comes from poorer individuals; iii) if, in Vancouver, housing regulations, such as zoning, were less stringent, welfare in the long-run would increase by approximately 5% in Vancouver and 0.57% in Canada, overall.