Various practices of grading, and some of their advantages and disadvantages are discussed. The first chapter presents the purposes and meaning of letter grades, and offers some guidelines: letter grades should reflect academic achievement, rather than affective behavior or personality; they should be closely related to course objectives and learning outcomes, and should be determined by a wide range of appropriate information sources. The second chapter describes several procedures for assigning letter grades, based upon: (1) percentage scores predetermined before testing; (2) group performance--normal curve using percentiles or standard deviations; (3) natural breaks, or gaps between scores; (4) group ability (score frequency); (5) growth or improvement; (6) performance contracts; and (7) criterion-referenced grading. Suggestions for combining evidence of achievement involve assigning relative weights to different data components, or combining components, using standard deviations or stanines. Ways to insure comparability of grades across teachers and subject areas include publication and discussion of letter grade distributions, use of an anchor test, and graphing of all assigned grades. Twenty-three discussion questions are included, as well as instructions for converting a set of scores into stanines, a 28-item glossary, and an 11-item bibliography. (GDC)