Economists are faced with the continual need for reliable estimates of farmers' aggregate supply response. They search accordingly for procedures that will reduce the computational burden and strengthen the reliability of these estimates. The author describes how the "variable resource programming" technique can be used to help alleviate a major problem encountered when "representative" farms are analyzed to generate aggregate information. This is the problem of determining the proper kind and number of farms to include in the analysis. The article also illustrates how variable resource programming can contribute a wealth of information pertaining to other production research problems, such as determining the optimum combination of enterprises for specific farm resource situations.