This study compared different stakeholders' perceptions of the validity of various indicators of student learning used to judge the quality of schools and individual student's academic performance. Data are based on questionnaire responses of 314 educators from school districts in three states that have implemented comprehensive statewide assessment programs that include high-stakes consequences both for educators and for students. MANOVA results showed significant differences between school administrators and teachers, with administrators favoring the validity and trustworthiness of nationally-normed standardized assessments, state assessments, and district assessments, while teachers favored classroom observations, classroom assessments, homework completion and quality, class participation, and behavior. The implications of these differences for reform initiatives are discussed, particularly with regard to teachers' motivation to improve results.