As part of an extensive research program, the Agricultural Estimates Division of Agricultural Marketing Service is investigating objective methods for estimating and forecasting corn yields. This paper is concerned only with one question: To what extent can dif ferences in estimates of yield per acre, derived from weighing small samples of the crop just before harvest, be reconciled with yields reported by farmers and the official yield estimates derived from such reports? The present status of information on that question is given here without distracting attention from the main issue by including a mass of technical statistical detail. it should be emphasized, however, that the materials contained in this article represent only the preliminary findings of this particular research project, and that final conclusions with respect to the validity of official corn-yield estimates, compared with those obtained by other methods, cannot be made until the research program in this area has been completed and evaluated.