Abstract Cognitive dissonance theory, as originally set out by Festinger (), described dissonance as an intraindividual phenomenon in a social context. Much of the research on dissonance has focused on the intraindividual aspect of dissonance. The limited research that has looked at cognitive dissonance in groups has done so from a range of different perspectives. These perspectives seem to result in contradictory predictions about the role of social information on the arousal and reduction of cognitive dissonance, despite generally sharing a model of the social self based on, or consistent with, social identity theory (Tajfel, ). Thinking about how these group‐based models of cognitive dissonance fit together may better illuminate the nature of dissonance and also suggest productive avenues for research to integrate these various perspectives on dissonance.