The purpose of the paper is to test the long-run steady state of growth factors hypothesized to influence U.S. manufacturing investment flows. These factors include agglomeration, market structure, labor, infrastructure, and fiscal policy. Spatial cross-regressive and spatial Durbin models are used to measure the spatial interaction of investment flows. Spatial spillovers are found to be of a competitive nature at the state level, implying that a factor which attracts more investment to a particular state is associated with lower investments in neighboring states. Investment flows to states with higher market demand, more productive labor, and more localized agglomeration of manufacturing activity.