This paper examines issues in gathering data for purposes of making accountability decisions for students with disabilities. The first section describes five alternative methods being used: (1) administration of norm-referenced tests; (2) focus on results or outcomes; (3) Individualized Education Program aggregation; (4) secondary analysis of extant data; and (5) establishment of an accreditation program. The second section looks at the student population and accountability decision-making practices in representative states (Illinois, Iowa, Kentucky, Maryland, Michigan, and West Virginia). Considered in the third section are the kinds of issues State Education Agency (SEA) personnel encounter and the barriers they run into as they attempt to document educational results--for students in general and especially for those with disabilities. These include agreeing on outcomes and standards, inclusion of students with disabilities in assessments, deciding on how and what to report, determining the role of the state, deciding on sanctions and rewards, and overcoming fear of accountability. The fourth section identifies the kinds of data SEA and national education policy personnel feel are needed to demonstrate that education works for students with disabilities. Section 5 describes ways in which SEA personnel have been successful in overcoming barriers. The paper closes by recommending promising practices for making accountability decisions. (Contains 10 references.) (DB)