The paper reviews the literature on the use of guidelines to identify handicapped children, especially learning-disabled (LD) children, for special programs. Dangers of misidentification of learning-disabled children through rigid use of discrepancy formulas are reported. The paper reviews studies which support: the use of Bannatyne's four factor recategorization model for learning disability identification; the use of diagnostic tests in conjunction with familial and cultural information to distinguish between LD and non-LD children; and the use of standard score and achievement discrepancies in determining eligibility for placement in LD programs. The paper also calls for consideration of the student's attention characteristics, methods of organizing and processing information, and the presence of specific behavior deficits. (11 references) (DB)