Mentoring programs at universities have become common because of the perceived benefit to student persistence and retention. Evaluation of the effectiveness of these programs has not kept pace, primarily due to the following three problematic issues: (1) lack of theoretical guidance, (2) lack of an operational definition of mentoring, and (3) lack of methodological rigor. This article describes the evolution of a regional Faculty-to-Student Mentoring program into a statewide program, and how it addressed each of these three problematic issues. Using logic modeling, the intimate connections between theory, operational definitions, and sound methodology are made explicit, thereby addressing many of the shortcomings of previous mentoring programs. By addressing these shortcomings, universities can better evaluate if mentoring programs should be part of the overall strategic plan to help students be successful.