Background: The transition to undergraduate studies consistently aligns with reductions in exercise, as well as psychosocial and physical health. We explored predictive pathways between exercise self-regulatory efficacy (SRE), exercise intentions, and exercise participation with undergraduates in their first semester at college. Guided by efficacy generality principles, we also investigated the way in which students' college (i.e., study-related) SRE predicted their exercise SRE beliefs.Method: Data were collected from 226 undergraduates (M-age = 19.11, SD = 2.88) over three time points, in order to explore predictive relations between psychosocial and exercise variables whilst minimizing error variance. At time one, students reported their college SRE and baseline exercise SRE, along with ratings of college and exercise importance. A week later, they again reported their exercise SRE, as well as their exercise intentions, and another week later they completed self-report measures of exercise engagement.Results: Structural equation modeling with bootstrapped confidence intervals revealed that exercise SRE positively predicted exercise engagement, both directly and indirectly via an effect on intentions. High college SRE also aligned with high exercise SRE, even when accounting for previous levels of exercise SRE. However, residual-centered latent variable interaction analyses revealed a significant moderator effect, such that college SRE predicted exercise SRE most closely when these two endeavors were viewed as being similarly important to one another.Conclusions: Conceptually, these findings provide novel insight into generality between SRE beliefs, as well as direct and indirect pathways through which undergraduates' SRE beliefs predict exercise engagement. From a practical standpoint, targeting students' college SRE may help to buffer against reductions in exercise levels and maladaptive health outcomes. (c) 2012 Elsevier Ltd. All rights reserved.