Did the Secure Fence Act lead to a change in migration patterns? Yes, but the effect was small. Using a unique dataset that contains information on the origin and destination of (primarily unauthorized) Mexican migrants, we document that migration fell between Mexican municipalities and U.S. counties that were more affected by the wall (i.e. by geographical proximity). The magnitude of the effects suggests that the direct effect of the wall expansion was to reduce migration flows by 0.8%. The direct effect, however, does not account for other effects, such as whether migrants changed where they migrated to or whether wages changed in the destination. To account for these, we develop and estimate a general equilibrium spatial model. We find the total effect of the wall expansion was to reduce the number of Mexican citizens living in the United States by 0.6%, or roughly 82,650 people.