To investigate the possibilities of establishing a representative sample of farms, using a system of data collection wherein farmers would report monthly their farm business expenses and receipts over a 12-month period, the U.S. Department of Agriculture entered into a cooperative agreement with Michigan State University in 1956. The Cooperative Extension Service had already established a farm record-keeping project wherein volunteer Michigan farmers submitted monthly financial reports to the Agricultural Economics Department of the University, which in turn processed the data currently using punched card equipment. The general success of this project and the need for current and accurate farm data by Government agencies led to the decision that such a system be offered to a group of farms selected on a probability basis. The study reported in this article was concerned more with the technical problems involved in establishing and maintaining a representative sample reporting certain data monthly than it was with the data so collected. If the technical problems could be economically surmounted, further consideration could be given to establishing such a system over a wider area—or an entire State or region, or even nationally. In this paper, the author describes the method of data collection used and the type of information obtained, discusses the problems of establishing and maintaining the sample, and evaluates the system in terms of one year of experience and its future potentialities. This paper is approved for publication as Journal Article No. 2444 of the Michigan Agricultural Experiment Station. The assistance of Olan, D. Forker, graduate assistant to the author, Nathan M. Koff sky, Ernest W. Grove, Earl E. Houseman of Agricultural Marketing Service, U.S. Department of Agriculture, and Wylie D. Goodsell, Agricultural Research Service, U.S. Department of Agriculture is gratefully acknowledged.