The literature on item bias, i.e., the question of whether some items in tests favor one cultural group over another cultural group due to irrelevant factors, is reviewed and evaluated. All known references through 1981 are described including a large number of unpublished reports. Each method is described and the criticisms that have appeared in the literature are reported. Studies evaluating methods, comparing methods, and using the methods in practical applications are described and criticized. In the two decades of work on this topic, little has been accomplished that will result in reduction in the mean differences between cultural groups. The probability seems high that well designed and developed modern tests contain few items biased in the sense defined above. (Author)