Two experiments were conducted to assess the impact of status differentials on subgroup attitudes and behaviours. In Experiment 1, 73 math-science students were led to believe they had higher or lower status than humanities students. They then performed a non-interactive decision-making task during which they were categorized exclusively as a university student (superordinate condition), or as a university student and math-science student simultaneously (subgroups condition). Experiment 2 (N = 98) differed from Experiment 1 in that perceptions of relative subgroup status were measured rather than manipulated. Consistent with social identity theory, subgroup members tended to categorize themselves more at the superordinate (university) level the lower status they considered their subgroup to be. In Experiment 2, a series of interactions also emerged, showing that status and inter-subgroup bias were positively related when the participants had been categorized exclusively at the superordinate level. When superordinate and subgroup identities were activated simultaneously, perceptions of status had no effect on levels of bias. The results were interpreted in terms of participants' needs for identity enhancement and identity distinctiveness.