Ellison and Glaeser (1997) find that almost all industries in the US are somewhat localized, but that the degree of localization is slight. Industries thus do tend to cluster geographically to benefit from industry-specific spillovers and natural advantages. Ellison and Glaeser (1999) find that some simple natural advantages can explain about 20 percent of the degree of geographic concentration. Duranton and Overman (2008) find similar results for the UK and also uncover new evidence on co-localization of vertically linked industries. Building on these results we analyze the geographic concentration of mergers & acquisitions (M&As). Unlike greenfield investments, firms cannot freely determine where to locate their M&A activity as they are bounded by their choice of existing establishments. The question thus arises if the distribution of M&As is similarly, more, or less concentrated than the distribution of sector activity as such. Since we use the observed concentration of existing sector activity as our basis for comparison there is no need to further control for the effects of natural advantages, either observed or unobserved.