Professional Development Schools (PDSs) emerged in the USA in the mid-1980s as a promising approach to simultaneously improving teacher education and the quality of education students receive in kindergarten through grade 12 (public) schools. This new form of school-university partnership enabled teacher educators to design new models for pre-service teacher preparation and in-service education for experienced teachers. While teacher candidates were routinely placed one per school in traditional, isolated placements in elementary schools with sporadic visits from the college supervisor, the school-university partnership relationship enabled the development of a model where four to six students were clustered in an elementary school with a college supervisor on site one full day per week and input from experienced teachers. This study evaluates the implementation and efficacy of two models of teacher candidate placement. Teacher candidates in Group A were placed in traditional isolated placements and those in Group B were placed in clusters of 4-6 students in a Professional Development School. In order to investigate the effectiveness of the cluster model, a study was designed that incorporated a multi-site ethnographic approach. Findings suggest that the clustered placements far surpass the traditional isolated placements in creating reflective, interactive educators and providing an effective setting for in-service education for experienced teachers. (Contains 1 table.)