Situated learning perspectives assert that knowledge is inseparable from the contexts and activities in which it develops and identify “communities of practice” as a way to characterize learning (Wenger, 1998). Knowledge and skill are acquired when new members “move toward full participation in the sociocultural practices of a community” (Lave & Wenger, 1991, p.29). The purpose of this study was to examine the various communities of practice that were formed to support the mentoring process in a physical education teacher development project. Goals of the project were to assist inservice teachers to examine and reframe their assessment practices and to increase their students’ knowledge and behaviors around physical activity. Participants included twelve teachers (3 males; 9 females) from six schools (2 urban; 4 suburban), project mentors (3 college/university faculty and 3 K-12 physical education teachers) and a project research team (2 graduate students and 5 faculty). Data included mentor record forms, notes from phone and email contacts, field notes from teacher development sessions, and interviews with teachers, mentors and researchers. Interview transcripts were coded using open, axial, and selective coding (Strauss & Corbin, 1998). Results indicated that the project began with clearly defined stakeholder roles (e.g., teachers, mentors, researchers). Through the course of the project, however, a series of highly interconnected and reciprocal relationships arose, representing communities of practice of varying dimensions. Teacher, mentor and research groups represented micro communities of practice while interactions among these groups formed larger interconnected macro communities (e.g., mentor/teacher, mentor/researcher, and teacher/researcher). Within this project framework, these communities represented support structures as well as planned and unplanned learning opportunities for all of the stakeholders involved. While some of these communities of practice were anticipated, others were more unexpected and arose as interactive responses to but not direct results of the project’s design. Collectively, the various subgroups formed a larger community of practice to address the project’s goals. Success of the various communities in this study depended on the active participation of all constituents. Productive communities were able to communicate openly, share knowledge and negotiate effectively. Groups without active participation among members did not experience these same benefits. Overall project success hinged upon successfully sustaining the various interconnected communities of practice. Results confirm that development efforts can provide structures for situated learning to occur within a community of practice based on the meaningful, purposeful, and authentic learning activities they provide (Lave & Wenger, 1991).Keyword(s): professional development