Communal conflicts threaten political stability throughout the world. While conflicts such as those between Hutu and Tutsi or between Bosnian Serbs, Croats, and Muslims degenerated to genocide, others such as those between the Anglophone and Francophone Canadians or between Walloons and Flemings in Belgium have remained remarkably peaceful despite chronic tension. This difference in the level of hostility across cases of identity-based conflict presents a rich field of investigation for political psychology. In addition, increased mobility and intergroup marriage has generated further political challenges to the notion of self-determination. It is increasingly difficult to categorize people and to adapt politically to changing demands, as indicated by debate in several countries over identity categories in censuses. People are increasingly conscious of the fact that they belong to multiple politically salient groups that sometimes have conflicting goals. In light of these difficulties, I present a theoretic basis for studying this interaction. Political studies of communal conflict and psychological research on identity operate at different levels of analysis, but can be combined to produce an explanation of the dynamic nature of identity politics. After a brief review of research in these areas, I present a model to adapt psychological theories of multiple identification to the complex realm of political debate and legitimation. Although there is not space here for a full application of the combination, I use examples from Belgium and Canada to illustrate this perspective. This political psychological perspective can greatly contribute to our understanding of the dynamics of identity-based conflict and efforts to intervene in such conflicts.