The goal of the current research was to investigate the impact of peer comparisons on female participants' math performance when the peers varied on comparison relevance and gender. Specifically, the two studies presented here explored the question of whether male peers, who are seen as less relevant, can serve as peer role models (PRM) for female participants taking a math exam under stereotype threat conditions. Comparison relevance was manipulated by exposing participants to math-talented male and female peers who were from participants' school or from another school. The key finding from Study1 demonstrated that exposure to relevant male peers lead to lower math performance compared to less relevant male peers. Study 2 replicated this key finding using a different manipulation of relevance. Together the findings indicate that when male peers are viewed as less relevant (and thus less threatening) they can serve as PRMs for female participants in "stereotyped" math situations.