Youth mentoring can have a profound impact on the lives of high-risk youth. This study presents the Campus Corps program, a time-limited (12-week), structured mentoring program for high-risk youth (ages 11--18), and results from a quasi-experimental pilot evaluation. Baseline and post-intervention problem behavior data from 315 offending youth were used in multiple regression analyses. After accounting for baseline group differences, pre-intervention scores, and demographic covariates, Campus Corps participants (n = 187, 63.1% male) reported less engagement in problem behavior, lower acceptance of problem behavior, and a greater sense of autonomy from marijuana use post-intervention than participants in the comparison condition (n = 128, 66.4% male). Conversely, post-intervention group differences were not observed for peer refusal skills or autonomy from alcohol use. A description of the Campus Corps program design and supplemental preliminary findings contribute to the growing knowledge base of youth mentoring program designs and outcomes.