This paper reviews the literature on assessment bias issues in special education. While assessment instruments yielding a single IQ score are seen as useful components in a comprehensive multifactored assessment, and are the primary tool of diagnosis for mental retardation, they are found to be irrelevant when applied to students with learning disabilities (LD), as many students with LD have specific deficits in the areas most emphasized by IQ tests. Recent litigation in four cases of alleged assessment bias is reviewed, highlighting the history of controversy over racial differences in IQ and disproportionately high placements of students from racial minorities in special education, as well as low levels of minority placement in programs for gifted and talented students. The validity of using IQ tests for placement purposes arises as a major issue in these cases, as does the stigma associated with labelling. Principles of culture-balanced testing are explored, and while arguments in favor of completely eliminating IQ testing in schools are presented, surveys are reported to find broad-based support for IQ testing among psychologists and educational specialists. Alternatives to IQ testing are discussed, with an emphasis on multiple intelligences models and dynamic assessment. (Contains 67 references.) (PB)