This paper studies the effects of a landmark college expansion policy on wages in Russia. We construct a unique individual-level data set linking the supply of college campuses in an individual’s place of residence at age 17 with educational attainment and hourly wages in adulthood. Using parametric and semiparametric selection models, we demonstrate that the returns to college education in Russia are heterogeneous and depend on whether individuals live in areas with limited college availability. The cost-benefit analysis demonstrates that discounted net benefits of college education for students in constrained municipalities exceed those for students in large cities twofold.