A taxonomy was developed for classifying instructional materials in fourth grade mathematics, as well as the content of actual classroom instruction and of achievement tests. The taxonomy is defined by the intersection of three factors and results in 468 cells: (1) mode of presentation (for example, use of graphs, tables, story problems)--3 levels; (2) nature of the numbers (single digits, fractions, geometric figures, and so on)--13 levels; (3) arithmetic operations--12 levels. This system was used to classify the mathematics subtests of four widely-used standardized achievement tests: the Stanford Achievement Test (SAT); Iowa Test of Basic Skills (ITBS); Metropolitan Achievement Test (MAT); and California Test of Basic Skills (CTBS). Three raters classified the content of these four tests, with inter-rater reliability ranging from 60% to 90%. Results indicated sufficient variance across tests to suggest that some tests may have more content validity for a given instructional program than others. A total score on any mathematics test represents an aggregate of many content areas. Since the taxonomy was able to denote content differences among tests, it was concluded that it would also be useful for an analysis of instructional content. (GDC)