This paper considers whether highly concentrated industries are better represented in the political process, as Olson's Logic of Collective Action suggests, and, if they are, whether this is so for the reasons that the Logic claims. It begins with a review and critique of the quantitative literature that has largely tried and failed to substantiate Olson's view. The bulk of the paper consists of five longitudinal case studies of firms that dominate or have dominated industries: IBM, Intel, Microsoft, America Online, and Cisco. The cases suggest that there is merit to the Olsonian view, but that alone it does not constitute an adequate political theory of the concentrated industry or the dominant firm. Additional variables drawn from organizational and institutional theory need to be incorporated into such a theory, including variables that bear on the allocation of attention, threat perception, and information flow within dominant firms.