A spatial interaction methodology is developed for modeling flows in a hierarchical system. A competing and intervening destinations framework is employed to model and predict US state-to-state labor migration. This analysis is used to assess the importance of geographic variables in explaining variations in regional labor flows. Empirical findings suggest that US labor migration is largely explained by Newtonian and systemic forces—size, distance, locational accessibility, and intervening opportunities in a spatial hierarchy. It is also suggested that lagged migration or migrant stock is a product of the combined effect of these forces.