This paper examines three groups of respondents (teachers, students, and principals) commonly surveyed in climate studies. The study sought to discover whether or not principals, students, and teachers agreed in their assessments of their school's disciplinary climate. The National Education Longitudinal Study of 1988 (NELS: 88) was used to create a Likert scale comprised of identical survey questions for each of the three groups. The hypothesis was that if this common measure of school climate possessed adequate reliability, schools that were ranked highly by one group would be ranked highly by the other two. Data show little evidence to suggest that students, teachers, and principals agreed on their perceptions of their school's climate. Significant differences were found among the three groups regarding specific aspects of schools' disciplinary climate. For example, students were over four times more likely than principals to see conflict among students as a significant problem. Further, the three groups viewed discipline in substantially different ways. Principals tended to have more positive images of school discipline than did teachers, who in turn were more positive than students. Researchers and practitioners should exercise a level of caution when using the concept of school climate in research and school-improvement efforts. Six tables are included. Appendices contain correlations matrices and a list of items and item names used for each analysis. (Contains 32 references.) (LMI)