Many states now use large-scale testing programs as a major mechanism in their educational accountability efforts. With the mandated participation of students with disabilities in large-scale testing programs, it is possible to use large-scale testing results to examine special education efficacy at the school level. We used special and general education reading outcomes from 6 elementary schools in 1 school district to compare and contrast measures of a school's performance over time in serving children with disabilities. Three different methods were contrasted: (a) percentage of students in special education reaching grade-level proficiency in reading each year, (b) percentage of students in special education exceeding expected growth in reading each year, and (c) longitudinal growth and percentage of students reaching grade-level proficiency in 5th grade. We discuss the results in terms of deriving a measure of school special education performance that is both accurate and useful for school personnel.