This work analyses intergroup comparison choices made by intermediate-status group members. Seventy-six psychology students were categorized in an intermediate position with respect to other faculties. Stability was manipulated at three levels: stable, upwardly unstable, and downwardly unstable. Data on strength of comparison, comparison for enhancing, comparison for evaluation, and ingroup identification were collected. Results revealed that in the stable condition, participants were equally engaged in both upward and downward comparison. In the upwardly unstable condition, participants were more likely to compare themselves with the high-status group, whereas in the downwardly unstable condition, they were more likely to choose a downward comparison. In this latter condition, both downward comparison for enhancement and in-group identification were lower than in other conditions.