We use longitudinal micro data to estimate the urban density premium for U.S. establishments, controlling for observed establishment characteristics and dynamic establishment behavior. We find that a doubling of urban density increases the average earnings of establishments by between 6 and 10 percent. The result holds after controlling for endogeneity issues and with the use of alternative measures of density. We find strong evidence against accumulated knowledge spillovers over time at the establishment level—that is, the density premium is realized at birth and is constant over the life of establishments. We find little evidence that the endogenous entry or exit of establishments can account for any of the estimated density premium.