In light of evidence indicating that referral itself often predicts student placement, an expert system was designed to assist educators to reduce bias in the process of referring students with suspected disabilities. A preliminary review of the literature looks at teacher perceptions as a predictor of handicapping conditions, referral bias, and placement as a measure of handicapping conditions. The Referral Consultant expert system asks a series of questions which quickly assess, first, whether the student is an obvious candidate for special education referral, and then, if not, the student's academic achievement and learning problems, possible handicapping conditions to account for the learning problems, other difficulties possibly accounting for the learning problems, and whether pre-referral interventions have been attempted. The output is a report describing its conclusions as to whether the student should be referred for special education assessment. The system was evaluated in 18 elementary schools. Implementation of the system had no statistically significant impact on the percentage of students referred for special education, or the ratio of students placed to students assessed. Additionally, variables predicting referral for special education assessment were not consistent. Another report detailing the Referral Consultant logic, questions, and output is attached. (21 references) (DB)