The nature of five mentor-intern relationships was investigated throughout the 1987-88 school year in four different, formalized mentoring projects in upstate New York. Subjects consisted of pairs of more- and less-experienced teachers. Variables considered in selecting the pairs included age, experience, gender, race, content area, context, subject area, ratio of interns to mentors, and first- and second-year participation by the school district. In-depth case studies were produced from data obtained from interviews and observations. Selective characteristics of the five pairs, as well as noteworthy aspects of four different local projects and five relationships are presented in outline form. Four out of the five cases documented satisfactory mentor-intern relationships. Conclusions drawn from the study include: (1) a satisfactory relationship can be created in a formal program; (2) considerable flexibility in the program is advisable; (3) a satisfactory relationship is enhanced when the pair receive common release time; (4) mentor training is important; (5) mentors should be self-selected, flexible, approachable, experienced teachers who can work well with adults; (6) formalized relationships can be fortified by personal as well as professional dimensions; and (7) an early start produces positive feelings about the program and the relationship. (JD)