Using a geographical approach, I exploit the regional (city) variation in the proportion of Mexican households deciding to send a migrant to the U.S across two quinquennial periods to estimate the causal effect of emigration on the local price indices of eight distinct groups of goods and services. To overcome the endogeneity of the emigration decision, I employ an instrumental variables approach that relies on the deep historical roots and high persistence characteristic of Mexican migration. My results show that emigration had a significant negative effect across all price index product classifications. Moreover, they suggested that emigration is more likely to affect non-traded good items and services than tradable products since the magnitude of the negative effect rose when restricting the analysis to the non-tradable good components within each price index classification.