It is asserted in this paper that some standardized tests do not do a good job of what they claim to do, and for some testing purposes nonstandardized tests are more appropriate or efficient. Under present circumstances, group I.Q. tests should be abolished. They provide no useful information that cannot be gained from achievement tests. And what they do provide--an I.Q. number for the cum card and the teacher's head--retards the academic progress of many children. On the whole reading readiness tests and the approach of which they are a part do more harm than good. It would be wiser to begin formal reading instruction, as some schools do, by attempting to teach all children the same things, without prejudicing or predicting their success. The standardized achievement tests given in the elementary and secondary schools are of little or no value to competent teachers in appraising the work of individual students. On the whole, college admission tests have a good record and continue to perform a useful function. If criterion-referenced tests tell us whether students have mastered instructional objectives, are they not better than norm-referenced tests? The answer depends on the purpose of the test. (Author/JM)