The traditional methodology for calculating economic optimum rates (EORs) of nitrogen (N) fertilization for corn (Zea mays L.) requires fitting models to describe relationship between yields and N rates. However, different models fitted to the same yield response data can result in disagreements in EORs, and the disagreements can be large enough to have economic and environmental importance. The objective of this study was to explore an alternative method for calculating optimal rates without fitting models when analyzing a large number of yield response observations. We analyzed a sample of 54 trials for corn after corn and corn after soybean with seven N rates applied to plots in a randomized complete block design with three replications. We used discrete marginal analysis where marginal physical products (MPPs) of N fertilization expressed as yield responses per kg of N applied were calculated for each fertilizer increment and segmented lines were used to interpolate MPPs between N rates applied. Economically optimal rates were calculated for maximizing profit per unit of area and for receiving the desired rate of profit on the last unit of N applied. The results showed that MPPs tended to be near linearly related to N rates in the near-optimal range. Discrete marginal analysis and linear interpolation in the near-optimal range was shown to be a simple and reliable method for calculating optimal rates. The utility of the alternative method is also in ability to provide better visual effects of changes in price of N and grain on optimal rates.