While the effect of teachers’ unions on school districts continues to be debated, the research literature provides few definitive conclusions to guide these discussions. In this article, we examine the relationship between teachers’ union contracts and school district efficiency. We define efficiency as the ratio of short-run productivity (student performance on standardized exams) to expenditures. We estimate a series of school district fixed effect models using measures of district collective bargaining agreement (CBA) restrictiveness tied to longitudinal outcomes. We find that CBA restrictiveness is positively associated with expenditures on students, instruction, instruction support services, and teacher and administrator salaries over time. We find no significant relationship between CBA restrictiveness and student achievement. Finally, we find a negative relationship between CBA restrictiveness and district efficiency. Given the small magnitude of our effect sizes, we conclude that weakening union rights may not produce large gains in efficiency and may come at substantial political costs.