Test-based accountability systems are now a central feature of U.S. education policy. Accountability systems are implemented as a way of improving student outcomes through new, highly visible incentives. This paper characterizes the incentives of different state accountability systems and reviews the existing evidence about outcomes. The question central to this paper is how different accountability measures reflect the quality and performance of schools and whether different accountability systems should be expected to generate improvements in student outcomes. The differences across states support a comparative analysis of the structure of the systems and the relationship between structure and performance of the systems over time. The paper discusses the theory behind and analyzes actual practice of cross-sectional and achievement-gain accountability systems. In discussing implications for policy and research, the paper states that more extensive and focused analysis is needed before strong statements can be made about the effectiveness of accountability for raising student performance. Included are 37 footnotes, most of which contain references. (WFA)