The sampling of test items found within the Science portion of selected standardized achievement test batteries is examined. The lack of sensitivity of such items to the curriculum changes often attempted in Earth Science is demonstrated. The following five standardized achievement test batteries commonly administered to pupils in grades 7-9 were selected for use in this study: Comprehensive Tests of Basic Skills, Metropolitan Achievement Tests, SRA Achievement Tests, Sequential Tests of Educational Progress (STEP), and Stanford Achievement Tests. Individual test items were classified into the general categories of Earth Science, Life Science, and Physical Science. Test items within general categories were further subdivided into content areas. With the exception of STEP, the three sciences were not represented within a given achievement test battery by equal numbers of test items. In addition, some of the achievement test batteries omitted or undersampled items in one or more of the content areas within the Earth Sciences. It is imperative to convince decision makers that standardized achievement test batteries should not be used to evaluate curriculum changes. (RL)