Staging events with a large number of participants is a central means by which collective action movements exercise power. Creating broad coalitions that cut across movement boundaries is one way to mobilize these large numbers. In spite of this fact, most studies of social movement coalitions focus on individual movements, analyzing them in isolation. This article explores the conditions under which organizations form alliances across movement boundaries, and examines whether these cross-movement coalition events are facilitated by the same factors that inspire coalition activity among organizations active within a single movement. I use event history methods to analyze data on 2,644 left-wing protest events that occurred on college campuses between 1930 and 1990. I nd several differences between the factors that facilitate cross-movement and within-movement coalition events. The availability of resources is important to within-movement coalition events but not to cross-movement coalition formation. Local threats inspire within-movement coalition events, while larger threats that affect multiple constituencies or broadly de ned identities inspire crossmovement coalition formation. The activity of multi-issue movement organizations is associated with higher levels of all forms of protest, including single and cross-movement coalition events. This research contributes to social movement theory by demonstrating that political threats sometimes inspire protest, and that organizational goals in uence strategic action. Throughout the 20th century, students frequently organized broad coalitions to protest against the policies of political and corporate elites. For example, in February 1987, when the chairman of Coors Brewing Company visited campus, 200 Harvard students protested (Landau 1987:1). Chanting "Coors, Coors, no way, racist, sexist, anti-gay," the students expressed their displeasure with the company's employment policies. The Democratic Socialists of America coordinated the event, which involved a broad coalition of student groups including the South Africa Solidarity Committee, the Harvard/Radcliffe Gay and Lesbian Alliance, the Committee on Central America, and a labor union. Coalition events such as this one raise a number of interesting questions that social movement scholarship has not examined. What factors facilitate collaboration across movement boundaries? What conditions inspire crossmovement coalition events rather than within-movement events? Research shows that social movement organizations that work in coalition with other groups are more likely to achieve success (Gamson 1990; Steedly and Foley 1979). Mobilizing large numbers of people and demonstrating widespread support for an issue is one of the few ways that social movements are able to exercise power (Koopmans 1993; Lipsky 1970; Tilly 1978). Creating broad coalitions that cut across movement boundaries is a central means by The author would like to thank