The standardization method of determining differential item functioning (DIF) is not acceptable for detecting biased items because of its methodological limitations. The main problems with the standardization index by N. J. Dorans and E. Kulick (1983) are: (1) it does not detect as DIF all items with large or aberrant group differences; (2) it considers as DIF items with average differences in passing the item equal to the average difference in ability between the two groups; and (3) in normal circumstances, any item will be considered as DIF if the difference in passing the item at each ability level is 8% beyond the difference in ability level of the groups measured by the percentage answering all items correctly. Applying this procedure to Scholastic Aptitude Test data for White, Black, and Hispanic students demonstrates its limitations. Several practical problems are also found when the procedure is applied to a licensure test. A practical method of detecting biased items is to use the differences in percentages passing the item among two groups. This simple method can reduce racial differences in test scores, as illustrated by its application to the results of an actual licensure examination that was administered to 343 applicants. Two tables present data on both techniques. (SLD)