This paper discusses the relationship of the content of nationally standardized and normed achievement tests and that of local school curricula and the effect that relationship has on the meanings and uses of the test scores. The following questions are considered: (1) whether tests have to match what is taught to be useful; (2) whether it is fair to teachers and students to use tests that include material not taught; (3) whether it is fair to use tests that do not include material that is taught; (4) whether local tests should be used instead of national tests; (5) whether national norms are needed; (6) whether some testing problems can be solved by the use of an item bank; and (7) whether or not it is possible to create a test that is tailor-made (customized) for local curricula and still obtain national norms by embedding nationally calibrated items. Five general issues are considered: inferences that can be made from standardized test scores; the basis upon which standardized achievement tests are considered valid; whether standardized achievement tests measure curricula; what the new technologies can do; and what schools can do. The respective answers to the seven questions are as follows: (1) no; (2) yes, provided the material is part of the target domain; (3) yes; (4) only for purposes where norms are not needed or where local norms (internal comparisons) are sufficient; (5) for most uses of test scores, national norms are the most easily understood and most informative type of norms--however, for some uses such as checking on just what the teachers are teaching, criterion reference scores may suffice; and (6) and (7) yes, in that more than one test can be blended together and put in one booklet--and no, in that unless the local curriculum very closely represents the domain sampled by a nationally normed test, two different tests are needed. Unless a very long test is acceptable, losses in reliability, domain coverage, and validity will ensue. (RLC)