I study knowledge spillovers in an industry where firms are mobile and heterogeneous in their ability to adopt outside knowledge (absorptive capacity). I develop a static model of industry agglomeration where, in equilibrium, the force of attraction induced by spillovers is counteracted by the force of repulsion created by local competition. The model is applied to a sample of the US software firms. I estimate the structural parameters of the model and obtain the following results: (a) The data are consistent with highly localized knowledge spillovers; (b) The attraction force induced by spillovers creates a significant sorting pattern placing firms with higher absorptive capacity in more agglomerated counties; (c) Ignoring firm heterogeneity in absorptive capacity leads to substantially biased estimates of gains from spillovers in policy experiments.