This paper estimates the short- and long-run effects of universities on geographic clustering of economic activity, labor market composition and local productivity and presents evidence of local spillovers from universities. I treat the designation of land-grant universities in the 1860s as a natural experiment after controlling for the confounding factors with a combination of synthetic control methods and event-study analyses. Three key results are obtained. First, the designation increased local population density by 6 percent within 10 years and 45 percent in 80 years. Second, the designation did not change the relative size of local manufacturing sector. Third, the designation enhanced local manufacturing output per worker by $2136 (1840 dollars; 57 percent) in 80 years while the short-run effects were negligible. This positive effect on the productivity in non-education sectors suggests the existence of local spillovers from universities. Over an 80-year horizon, my results indicate that the increase in manufacturing productivity reflects both the impact of direct spillovers from universities and general agglomeration economies that arise from the increase in population.