Agglomeration research has investigated a key research question, i.e., why do firms in a specific industry co-locate geographically? In the agglomeration literature, it has been assumed that each firm has one business establishment in a cluster such that firms always co-locate with competitors. However, it is often observed that firms operate several business establishments in a cluster, so they co-locate not only with competitors (i.e., inter-firm agglomeration) but also with their own business establishments (i.e., intra-firm agglomeration). While inter-firm agglomeration is a counterpart to the traditional concept of agglomeration, intra-firm agglomeration is a new concept in agglomeration research. The separation between intra-firm and inter-firm agglomeration raises two research questions – 1) how does intra-firm agglomeration differ from inter-firm agglomeration? and 2) do firms decide their locations for intra-firm vs. inter-firm agglomeration differently? These questions actually extend the key question in agglomeration research into a new setting in which firms have several business establishments in a cluster. I proposed that firms can extract more benefits but neutralize more threats from agglomeration through intra-firm agglomeration than through inter-firm agglomeration. I further developed research hypotheses to test this argument in a research context in which multi-unit firms decide their new establishments’ distances to competitors and their other establishments at the same time. The hypotheses received empirical support in an empirical setting in which 10 large multi-unit hotel firms established new hotels in 20 U.S. cities, and several supplementary analyses show that these results are robust.