Findings from recent research about the relationship of teacher empowerment to other school reform objectives of interest, such as classroom practices or student academic performance, are mixed. This study investigates teacher empowerment in schools that have at least four years of experience with some form of decentralized or school-based management. Based on the assumption that participation in school decisionmaking can enhance teachers' commitment, expertise, and, ultimately, student achievement, we hypothesize a positive relationship between empowerment and student performance through the linkages of school organization for instruction and pedagogical quality. The data we use to examine empowerment are drawn from a sample of 24 restructuring elementary, middle, and high schools - 8 schools at each grade level. Most of the schools are urban, representing 16 states and 22 school districts. Data sources include teacher surveys, ratings of pedagogical quality, assessments of student academic performance, and case studies based on interviews and observations; the primary method of analysis is hierarchical linear modeling (HLM). The results suggest: (1) Overall, empowerment appears to be an important but not sufficient condition of obtaining real changes in teachers' ways of working and their instructional practices; (2) The effects of empowerment on classroom practice vary depending on the domain in which teacher influence is focused; (3) Teacher empowerment affects pedagogical quality and student academic performance indirectly through school organization for instruction.