Many organizations currently implement formal mentoring programs to assist newcomers in their adjustment to organizational norms, standards, and culture. However, very few empirically-established guidelines exist for how to effectively match mentors to proteges in formal programs sponsored by organizations. Typically, organizations attempt to match mentors to proteges based on similarity of goals/interests. However, prior research suggests that even mentors and proteges disagree with respect to their perceived similarity. Consequently, it should be difficult for a program administrator to determine which mentors and proteges are likely to be compatible. Recent research has found that proteges who perceived they had input into their match reported higher quality relationships. The present study extended this research by experimentally manipulating protege choice and by investigating potential mechanisms through which choice is expected to influence relationship success. Mentors were undergraduate juniors and seniors and proteges were first-semester freshmen randomly assigned to choose their own mentor or to be matched to a mentor by the program administrator. Participants then met online in a private chat room once per week for a period of four weeks (30 minute sessions). Results indicated that when proteges were given the opportunity to choose their own mentors, both mentors and proteges felt more similar to one another. Additionally, proteges had higher expectations for what they would get out of the relationship and were more proactive in soliciting guidance from their mentors. Finally, each instance of coded psychosocial support demonstrated by a mentor related more positively to proteges' ratings of the support they received if they were in the choice condition. In fact, the relationship between coded psychosocial support and protege ratings of psychosocial support was slightly negative for those who were assigned to a mentor by the researcher. Pre-meeting expectations were found to fully mediate this effect. Finally, protege-reported psychosocial support was positively associated with self-efficacy and negatively related stress after the fourth chat session. In summary, the results of this study strongly suggest that protege involvement in the match process can facilitate the quality of formal mentoring programs.