Heavy drinking in college remains a concerning issue due to its association with both health and social risks. While modelling contributes to college students’ alcohol use, little work has identified who might be most susceptible to modelling effects. Peterson, Morey, and Higgins (2005) found males high in extraversion were more susceptible than others to matching strangers’ drinking levels in a lab-based social drinking context. We sought to replicate and extend these findings by examining the impact of extraversion on social matching of alcohol consumption levels of a drinking buddy in college students’ real lives. First, a significant relationship between buddy and target drinking levels was predicted in dyadic drinking situations. Additionally, we hypothesized that target extraversion would positively moderate this buddy- target drinking levels relationship. Data from 149 college student targets (74% F) and their same-sex drinking buddy were collected through online questionnaires examining targets’ extraversion levels, and the drinking levels and social drinking context of both dyad members through a 30-day Timeline Followback measure. Linear mixed-effects modelling confirmed the study’s first social matching hypothesis, while also revealing that target extraversion positively moderated the relationship between buddy and target daily drinking levels in dyadic drinking contexts. Findings extend those of Peterson et al. (2005) to a real-world (vs. lab-based) context, modelling of a buddy’s (vs. stranger’s) drinking levels, and a sample including women (vs. all-male). Results provide novel information on extraversion’s contributions to modelling of alcohol use that may guide useful modifications to personality-based interventions for reducing college student heavy drinking.