Nitrogen response curves derived from experimental data are used with a profit maximising condition to obtain optimum N rates, yields and gross margin losses for a range of nitrogen and cereal prices. These results are used with a linear programming model of an arable farm to estimate changes in the optimal cropping allocation and hence the farm scale effects of relative price changes. Total nitrogen applied is found to have a limited response: a doubling of the N price reduces the total used by between 24% and 10% depending on the availability of low‐N break‐crops. Cereal price changes reduce profitability severely before having any significant effect on N