Background: Social comparison research usually demonstrates that students will have higher self-evaluation in downward comparison but lower self-evaluation in upward comparison. However, the existence of this contrast effect may depend on people's self-construal. The contrast effect may exist only for people with independent self-construal. For people with interdependent self-construal, the contrast effect may be attenuated. Aim: The study investigated the role of self-construal as a moderator of the social comparison effects in authentic classrooms. Sample: The participants were 96 Chinese seventh-grade students (41 male, 51 female and 4 unreported) from a secondary school in Hong Kong. Method: The experiment employed a 2 x 2 between-subjects design based on 2 levels of self-construal (independent, interdependent) and 2 levels of comparison standard (upward comparison, downward comparison). The dependent variable was students' self-evaluation. Results: A two-way ANOVA indicated a significant interaction between self-construal and comparison standard on self-evaluation. When the students' independent self-construal was activated, they reported higher self-evaluation in downward comparison but lower self-evaluation in upward comparison. However, such a contrast effect was attenuated when the students' interdependent self-construal was activated. They reported high self-evaluation in both upward and downward comparisons. Conclusions: The outcome of social comparison depends on whether independent or interdependent self-construal is salient in the classroom.