This research investigated whether classroom-based peer norms for achievement goals moderate friendship selection, maintenance and influence processes related to academic achievement in 46 Grade 5 and Grade 6 classrooms (N = 901, 58.7% Grade 5 students, 48.5% boys). A distinction was made between peer norms for mastery (i.e., developing competence) and performance (i.e., demonstrating competence) goals. Peer norms were measured in terms of popularity norms (the within-classroom correlation between student achievement goals and popularity) and descriptive norms (the class-level aggregated average achievement goals). As hypothesized, longitudinal social network analyses revealed that achievement goal popularity norms played a role in friendship processes, rather than achievement goal descriptive norms. Specifically, adolescents formed friendships with similarly achieving peers in classrooms with high performance goal popularity norms but not in classrooms with low performance goal popularity norms. Conversely, adolescents remained friends with similarly achieving peers in classrooms with low performance goal popularity norms but not in classrooms with high performance goal popularity norms. Furthermore, friendship influence on achievement took place in classrooms with high mastery goal popularity norms, but not in classrooms with low mastery goal popularity norms. This study indicates that friendship processes regarding achievement depend upon the extent to which certain achievement goals are made salient by virtue of their association with popularity in classrooms.