This article explores the affordances and risks of practicing friendship and mentorship as methodological approaches in two qualitative studies: (a) the mentor's study in a diverse 9th grade classroom and (b) the protege's subsequent study of teacher professional development in the same school. Friendship methodology, as theorized by Tillmann and others, is extended to include protection and mentoring. The effect of mentoring is demonstrated through examples of the former protege's own research. Explosive moments in each study demonstrate how research can be analyzed and the course of the research projects influenced within a friendship/mentorship context. Like friendship-as-method, mentorship as methodology can result in rich data, but there is also the potential for more transparent and rigorous data analysis when the researcher is a mentor because the mentor can model research skills for the protege-participant. Thus, mentorship as methodology socializes peers into the conventions of qualitative research. (Contains 7 notes.)