The study was designed as one component of a program evaluation in special education funded by the State Department of Education in Virginia. The purpose was to identify the characteristics of mildly handicapped children who had been identified as learning disabled, educable mentally retarded and emotionally disturbed in a small school district. The parameters of the study included demographic information, school-based characteristics and assessment data. Student data were collected from the confidential folders maintained by the school district which included all written reports required for assessment and placement. Demographic data were taken from cumulative folders maintained in the schools for the students and reports developed by the building principal. A regression formula was used with the assessment data to determine if learning disabled students exhibited a severe ability-achievement discrepancy statistically. The regression formula and grade equivalent comparisons were conducted with emotionally disturbed students to determine if the emotional disturbance was adversely affecting their educational performance as measured by standardized tests. A constant comparative method was used to analyze the minutes from eligibility committee meetings to determine the important features school-based teams used for their assignment of labels to handicapped children. Comparisons were made between the characteristics of the children identified as handicapped and the state and federal definitions for those handicapping conditions. A discriminant analysis was used to investigate the possibility of predicting which students considered eligible for special education services would be classified learning disabled or emotionally disturbed based on 10 variables. Interviews were conducted with the program evaluation stakeholders committee to solicit their feedback concerning the results of the study.