The initial year of university is often a sensitive period for new students. Commencing students may lack the necessary skills and resources to adapt to unfamiliar learning environments. One intervention demonstrating academic benefits is Peer Assisted Study Sessions (PASS). PASS is a structured peer led study group where students collectively share knowledge and solve course-related tasks. To date there has been limited empirical exploration into "how" PASS enhances student performance outcomes. To amend this gap, the current study used both a cross-sectional (n = 264) and a matched longitudinal (n = 76) survey design, combined with PASS attendance and course performance data, to investigate three psychological mechanisms that may mediate these effects: increased academic engagement, a positive student identity, and increased statistics self-efficacy. Sampling a first-year psychology cohort enrolled in an introductory statistics course, both cross-sectional and longitudinal analyses found a positive relationship between PASS attendance and academic performance. Furthermore, self-efficacy mediated the relationship between PASS attendance and student performance.