Drawing on data from a US study of nine mentors and mentees, including mentee scores on the Reasoning about Current Issues (RCI) Test, which offers a measure of cognitive complexity, the authors explore how differences in cognitive complexity were related to role expectations, conceptions of teaching problems, and the use of evidence for justifying beliefs. Growing out of the Reflective Judgment Model developed by King and Kitchner [(2002). "The reflective judgment model: Twenty years of research on epistemic cognition." In B. Hofer, & P. Pintrich (Eds.), "Personal epistemology: The psychology of beliefs about knowledge and knowing" (pp. 37-61). Mahwah, NJ: Lawrence Erlbaum Associates] the RCI defines cognitive complexity in terms of how individuals reason--make judgments and use evidence--about ill-structured, controversial, problems. The authors argue that differences in how individuals reason when problem solving may help explain some aspects of how relationships between mentees and mentors form and suggest that greater attention needs to be given to cognitive complexity when designing induction and mentor programs.