The introduction of numerous classifications for sandstone during the past 25 years has created serious problems of nomenclature. Presently, identical names can be given to rocks having quite different properties, and, conversely, quite different names are given to rocks having nearly identical properties. It is suggested here that numerical classification of sandstone can be substituted for verbal classification to alleviate such problems of duplicated nomenclature and conflicting terminology. Numbers do not have the descriptive appeal of words, but they can be applied with far greater precision and objectivity and much less confusion. The simple, numerical classification for sandstone introduced in this paper is based on mineralogic and textural maturity, which are seemingly the most important criteria for classification. Relative mineral abundance is chiefly a function of relative mineral stability, and this relationship can be used to separate the sand-size components of sandstones into three major groups; siliceous resistates, feldspars, and labile grains. The new method combines the relative percentages of these components and matrix into a workable, numerical classification. As an example, a Class 6-2B sandstone is immediately defined as a sandstone containing 10-20 percent matrix and 80-90 percent sand-size grains; the sand-size grains consist of 60-69 percent siliceous resistates, 20-29 percent feldspars, and 2-20 percent labile grains.