The authors argue that personnel who select standardized achievement tests have been led to believe that the major achievement test batteries differ very little in terms of the topics they test; but that the content covered by these major tests is different, and that such differences have consequences for instructional content. To test this argument, four fourth grade mathematics achievement tests were examined: the Stanford Achievement Test (SAT), Iowa Tests of Basic Skills (ITBS), Metropolitan Achievement Tests (MAT), and the Comprehensive Tests of Basic Skills (CTBS). Test content was compared to topics which a teacher might cover. In the "operations" category, the tests corresponded quite closely in the percentages of items involving certain adding, subtraction, and division tasks. However, important differences were found in other categories and the number of items covering addition, grouping, alternative number systems, and percentages varied considerably between tests. Evidence of high internal consistency or test reliability does not rule out the effects of item-treatment-interactions. It is suggested that the problem of teachers teaching only the content covered by the test is not as great as might be thought. The authors recommend that tests be selected consistently and that their content be valid for the curriculum content. (Author/GDC)