Because of the limitations of the traditional approaches to classifying a child's characteristics, a number of authors have proposed to describe children on the basis of a defined set of functional skills or abilities rather than by etiological or deficit category (Bailey et al. (1993); Holt (1957); Linden (1963)). Building on this earlier work, Simeonsson and Bailey (1988) developed the ABILITIES Index (AI), an index designed to assess functional characteristics rather than skills. Although the ABILITIES Index was not originally designed to link student characteristics to the patterns of expenditure variations, the size of the SEEP (Special Education Expenditure Project) student sample offers a unique opportunity to apply the AI classification scheme to explore how such a tool might be used to explain the patterns of variations in expenditures on students with disabilities. The purpose of this paper is to compare the variance in expenditures explained by the IDEA (Individuals with Disabilities Education Act) disability categories with the dimensions delineated in the ABILITIES Index. The report demonstrates that the AI and IDEA disability categories are useful both independently and together in explaining the variation in expenditures on students with disabilities, while controlling for other student background and community/regional variables. The analysis that follows includes three stages. The first stage determines the percentage of the variation in spending explained by the student's primary disability (from among the 13 IDEA disability categories). The second phase explores whether a particular child has a second, third, or more disabilities, and explores the extent to which these disabilities contribute to expenditure variations. The third phase of the analysis examines an alternative approach to classifying students. A bibliography is included. (Contains 10 exhibits and 8 footnotes.)