All of the models discussed so far have their logical foundations in the individual firm and its entrepreneurial decision-making process. This is especially true if there is to be any attempt to explain variations in efficiency as due to managerial characteristics and ability. Aggregation beyond the firm level is likely to obscure any such impact. However, no data set containing individual firms was available in a time series of cross section observations in which farm operator characteristics were also reported .5 The data set used here is an 8 X 48 matrix, where each of the forty-eight contiguous states is considered a “farm firm” and the observations are over the eight-year interval 1960-67. What will be estimated, then, is an aggregate U.S. agricultural production function, similar in philosophy and execution to those estimated by Griliches (1963a, 19636, 1964). The results presented here will extend those of Griliches in two directions: (1) a frontier production function will be estimated, and (2) AC techniques will be used to remove management bias from the average production function