We develop a unified theoreticaland empirica lframework to study the impactof trade shocks on local labor markets.We start by explicitly characterizing, in a class of trade and geography models, a structural relationship between trade and labor market outcomes at the regional level.We show that both labor supply and agglomeration elasticities are central in determining this relationship. To identify these key elasticities, we propose a new empirical methodology that uses as an instrument the impact of exogenous trade shocks on changes of the endogenous variable predicted by our general equilibrium model. This methodology yields the most efficient estimator of the structural elasticities, i.e. a Model-implied Optimal IV (MOIV). We then appl your methodology to evaluate the aggregate impact of trade shocks affecting regional labor markets in the U.S. Join with Rodrigo Adao (Chicago Booth, United States) and Federico Esposito (Tufts University, United States).