Agricultural price analysis was one of the hard cores around which all the agricultural economics of the 1920's and early 1930's were built. Since then, in all too many cases the working economists have been too busily engaged in current operations to set down their appraisals of price-making forces in any formal way. Many have drifted from recognized statistical methods to a shorter-run, almost wholly intuitive, "market feel" approach. Some of the theoretical or teaching economists, especially the mathematically trained group, have gone in the opposite direction, stressing models, structural equations, and the substitution of symbols for statistics. In one sense this article returns to an earlier tradition, once again substituting statistical values for symbols, and at the same time formally setting down both the methods and the results in such a way that they can be checked, in terms of both theory and experience. But Fox has gone beyond the earlier tradition in a number of respects. Commodities accounting for a large proportion of farm income are treated in a consistent manner. The marketing system is recognized as a separate entity standing between consumer demand at retail prices and that of processors and dealers at the farm or local level. the statistical methods used are relatively simple, but they have been chosen after careful consideration of the theories and more complex equation forms advanced by mathematical economists and econometricians. Suggestions are offered as to means of reconciling both family-budget and time-series information relating to the demand for food. The more technical part of the article is preceded by a discussion of factors affecting the general level of farm income and the demand for farm products as a group.-O.V. Wells.