This is a study of distributed leadership in the context of elementary schools' adoption of comprehensive school reforms (CSR). Most CSRs are designed to configure school leadership by defining formal roles, and we hypothesized that such programs activate those roles by defining expectations for and socializing (e.g., through professional development) role incumbents. Configuration and activation were further hypothesized to influence the performance of leadership functions in schools. Using data from a study of three of the most widely adopted CSR models, support was found for the configuration and activation hypotheses. Leadership configuration in CSR schools differed from that of nonCSR schools in part because of the addition of model-specific roles. Model participation was also related to the performance of leadership functions as principals in CSR schools and CSR-related role incumbents were found to provide significant amounts of instructional leadership. Further support for the activation hypothesis is suggested by positive relationships between leaders' professional development experiences and their performance of instructional leadership. (Contains 7 tables, 1 figure and 8 notes.)