Social movement scholars have contended that considerable research on protest policing has been done, but research testing multiple theories in recent decades is lacking. To resolve this gap in the literature, this study integrates major paradigms in repression research and theories of policing styles around a dramaturgical approach to collective action, identifying factors influencing violence at social protests in the United States from 2006-2009. Conceiving of social protest as a form of political and symbolic action, I maintain that social actors and the qualities of their actions and immediate environment importantly influence a protest event's characteristics and outcomes. Specifically, I code for three violent outcomes—arrests, police violence, and any violence—and one measure of threat—police presence. I identify four components of the protest event which influence these outcomes—actors (e.g., authorities, protesters, and counterprotesters), enemies (e.g., the target of protesters' claims), the stage (e.g., qualities of place and space where a protest occurs), and protest performance (e.g., protest size and specific tactics employed by actors). Thus, this research focuses on how qualities of police, protester, and counterprotester performances intersect to influence violence at protest events. Data for this project were collected from multiple sources from 2006-2009. Information on protest events was collected by content coding of newspaper articles in the Los Angeles Times and New York Times. Information on community policing styles was derived from lists of funding for agencies participating in the U.S. Department of Justice's Community Oriented Policing Services (COPS) program. In some instances the results of this study show that certain characteristics leading to police presence and violence at social protests in the U.S. persist from research conducted on earlier decades—presence of African Americans or counterprotesters, protester use of "more confrontational" tactics and/or multiple tactics, and the damaging of property by protesters or counterprotesters. However, my findings also contradict previous studies, because I find that: (1) larger protests are less likely to be policed or result in violence; (2) social and cultural targets are more likely predictors of policing and violence rather than government or economic ones; and, (3) specific social movement families and tactical types influence protest event outcomes differently. I also found that community policing styles had no effect on protest event policing. These findings are important because they show how a protest event's symbolic nature influences policing and violent outcomes. / Ph. D.