Community-based participatory research has shown great promise as a mutually engaging and respectful way to partner contemporary biomedical knowledge with the lived experience, wisdom, and customs of American Indian people. Designed and implemented through this approach, our Family Education Diabetes Series (FEDS) has evidenced pilot and longitudinal physiological data supporting its effectiveness. However, the multifaceted nature of the program makes it difficult to know which factors are responsible for its success. This difficulty hinders efforts to improve the FEDS and/or inform others' work to advance similar projects. In this study, we conducted a qualitative investigation using talking circles to explore participants' views about what elements of the FEDS are most salient. Our findings suggest that social support and group-oriented sequences hold the most value. We conclude that an emphasis on these processes (instead of program content per se) is most indicated in effecting behavior change and facilitating ongoing disease management.