Despite extensive research on political activity on the part of corporations, clear and consistent findings remain elusive. We identify three reasons for this failure. First, most of the empirical literature on corporate political activity simply studies the wrong phenomena by examining political action committees rather than lobbying more generally. Second, the literature studies an excessively narrow sample of organizations that might engage in lobbying, focusing almost always on extremely large corporations, which inevitably attenuates variance on many of the variables hypothesized to influence engagement in political activity. And third, prior work is rarely attentive to the diversity of corporate activities, narrowly conceptualizing vital aspects of the business context that might influence decisions to engage in political activity. Based on this critique, we develop and test new models of corporate political activity, finding that the diversity of the economic context within which firms work and firm size matter a great deal, if in ways somewhat different from those reported in prior work.