There are many potential sources of test bias besides that of the particular item content within the test. The other potential sources, designated here as program and utilization must also be encompassed in any thorough and effective program to increase the accuracy of assessment for members of ethnic minorities. As usual, the research findings are emerging much more slowly than we would like, but that is the nature of careful research. Meanwhile, our failure to find bias from those sources that are most often identified, such as test content or predictive validities, must not be used to justify an abandonment of the search. The research efforts must encompass these other possible sources of inaccuracy, too; for that matter, they should be continuous, serving a monitoring function of these possibilities. But meanwhile, we can be aware of them and of the things that can be done to increase measurement accuracy using research evidence or lacking that, just good judgment and sensitivity. (For related documents, see TM 003 063-064, TM 002 964.) (Author/KM)