We tested some implications of Wills' (1981) downward comparison interpretation of ingroup bias in the minimal intergroup paradigm. Based on a self-enhancement interpretation of ingroup bias, we predicted that subjects who expected to succeed on a task for dispositional reasons and subsequently failed would be most threatened by the feedback and hence, would engage in downward social comparison strategies. The results did not support the self-enhancement interpretation, but a number of interesting findings emerged. First, downward social comparison involving favorable comparisons of the ingroup relative to the outgroup was pervasive and not mediated by self-esteem. Second, ingroup bias was greatest when individuals' outcomes were consistent with their expectations; ingroup bias was mitigated when subjects received feedback that was inconsistent with their expectations. Third, although low self-esteem subjects rated members of the outgroup more negatively than did high self-esteem subjects, high self-esteem subjects engaged in more downward social comparison by enhancing the self relative to both members of the outgroup and their own ingroup. Finally, self-enhancement strategies were affected by performance expectations, attributions, and chronic self-esteem: People who expected to perform well because of stable, dispositional reasons and who were high in self-esteem showed the greatest tendency to engage in self-enhancing comparisons with others. This was true regardless of whether subjects ultimately succeeded or failed on the important task and regardless of whether the comparison others were members of the outgroup or the ingroup.