--- - |2 Abstract Mentoring is modestly effective for youth with a chronic illness or physical disability; however, program effectiveness may be enhanced when mentors and mentees are matched on shared interests and experiences. To test this hypothesis, mentees were randomly assigned to having a mentor with or without visual impairment (VI). Results showed that mentors without VI were younger and more likely to work or be educated in a helping profession and less likely to have a fixed mindset and extremely high positive expectations than mentors with VI. The main analyses on match outcomes showed that mentors with VI had significantly fewer and shorter match meetings, had a weaker relationship with their mentees, and were more likely to end their match prematurely. Mentor age, helping profession background, and fixed mindset were confounds in several analyses and reduced the significance of the relationship between VI group and match meeting quantity. The only relationship that remained significant controlling for covariates showed that matches including a mentor with VI were significantly more likely to end in premature closure than matches including a mentor without VI. Implications of the findings for future research and program practices related to matching were discussed. - Journal of Community Psychology, EarlyView.