Researchers have long postulated the existence of a big-fish-little-pond effect (BFLPE) in which a student's relative academic standing in class or school affects his or her academic self-concept. Few studies, however, use causal research designs to identify whether the BFLPE exists and whether it is generalizable across a wide variety of contexts. The goal of our study is to provide causal estimates of the BFLPE and examine whether the estimates differ by gender and national context. To fulfill our goal, we analyze cross-national TIMSS 2011 data using a cross-subject student-fixed effects model. Our results provide the strongest evidence to date that a sizeable BFLPE exists in STEM subjects regardless of gender and national context.