Among 53,334 urban Colorado establishments, geographic information systems (GIS) techniques identify those that are and are not in enterprise zones (EZs). EZs have no effect on payroll per worker. Therefore, subsidy incidence is not on labor. Urban EZs do not increase employment per establishment, implying that subsidies induce net substitution effects for capital that counteract scale effects on labor. Rural EZs increase employment in the smallest of 13,278 establishments, implying that capital is less substitutable for labor with rural production functions and subsidy mixes. Employment effects differ across EZs. Equilibrium incidence of subsidies is probably on immobile factors such as commercial real estate.