Abstract: Abstract: Highest yield and highest ratio of yield to cost are 2 important parameters in agricultural land quality grade. The procedures of agricultural land quality classification (GBT28407-2012) points out that the highest yield should be investigated and determined by the land administrative departments of the provinces according to the specified crop species in standard farming system, and the maximum highest ratio of yield to cost for the specified crop species in the partitions of provinces should be determined by the relevant experts. Due to the lack of scientific basis, the determination of highest yield and highest ratio of yield to cost was subjective and random, which affects the comparability, scientificity and rationality of agricultural land quality grade results. In this paper, we applied wheat input-output survey data of farmland in Henan Province, and divided the data into 2 parts, one for research and the other for verifying the research results, after eliminating abnormal sample data. Based on the theory of land diminishing returns, the wheat per unit yield and per unit input of the agricultural land quality grade were used as the total yield and the total input of the land diminishing returns curve respectively, and accordingly the average yield had the same meaning as the production-cost index. Then the relationships among unit yield, production-cost index and marginal yield of diminishing returns curves were used to determine the highest yield and the highest ratio of yield to cost. The paper concluded that the theoretical conditions for achieving the highest yield and the highest ratio of yield to cost were as follows: In the complete marginal yield diminishing returns curves, the intersection point of the marginal yield curve and the X axis was the highest yield; in the incomplete marginal yield diminishing returns curves, the right end of the yield curve was the highest yield; the intersection point of the production-cost index curve and the marginal yield curves was the highest ratio of yield to cost. A correction method for the abnormal production-cost index was also proposed in this paper. In the incomplete marginal yield diminishing returns curves we obtained, the highest yield of wheat was 8 850 kg/hm2, 2.16% lower than the highest yield of 9 041 kg/hm2 obtained from the yield curve trend extrapolation theory, showing that they were very close. Compared with the original highest yield of 9 400 kg/hm2, the space transition characteristics of wheat land use coefficient of equivalent area obtained by using the new highest yield calculation method could be showed better, the spatial differentiation of the coefficient was more in line with the actual production, and the correlation degree between new wheat utilization class index calculated at provincial level and validation sample yield was higher; compared with the original highest ratio of yield to cost of 3.70 kg/yuan, the highest ratio of yield to cost determined in this study was 2.00 kg/yuan according to the wheat price at the time, which conformed more to the local wheat gains at the time. The results showed that calculation method of the highest yield and the highest ratio of yield to cost was scientific and feasible. The determined highest yield had corresponding funding support and the determined highest ratio of yield to cost had economic significance and practical value. The results verified that "high land yield doesn't always mean high production-cost index". The research provides scientific basis for calculating the abnormal production-cost index, enriches the connotation of the utilization coefficient and the economic coefficient, and improves the compatibility, rationality and scientificity of the agricultural land utilization grade and the economy grade results.