EDUCATION may have productive value in agriculture both because it may enable a farmer to produce more output with given inputs of other factors and because it may help him in obtaining and using information for managerial decisions on the purchasing of factor inputs and the choice of products to produce.' This study deals with the second of these two roles: it attempts to measure the influence of education on the cost-efficiency with which farmers combine various broadly defined inputs. The data refer to counties in Indiana, Illinois, Iowa and Missouri and were taken from the 1959 and 1964 U.S. Censuses of Agriculture. Section II sets out the basic model and section III describes the data, the estimation methods and the results obtained. Section 5 of the Appendix gives further details of the data used. The most important finding is that, at a given scale of production, farmers with above average levels of education managed to operate significantly nearer than average to the theoretically estimated point of minimum cost. This finding is consistent with the results of a study by Huffman (1972).