A gronomy J our n al • Volume 101, I s sue 2 • 2 0 0 9 269 ABSTRACT Marked spatial and temporal variability in yield response to N fertilizer observed in individual yield response trials creates a high degree of uncertainty when estimating economic optimum rates (EORs) of N for a group of trials and when extrapolating these rates from one location to another. A survey was conducted to characterize and classify variability in yield response to N on subfi eld and fi eld scales. Fertilizer N was applied at fi ve rates (56, 84, 112, 140, and 168 kg N ha -1 ) in many (6-12) replicated strips within three 18-to 24-ha no-till fi elds during two corn (Zea mays L.) growing seasons. Yield responses or yield diff erences between two adjacent strips were measured in 22 to 25 grid cells ha -1 within each fi eld. Cumulative probability distributions (CPDs) were used to estimate the probability that a given N rate produces a yield response less or equal to a specifi ed quantity. Th e yield responses were classifi ed into potential categories with diff erent N fertilizer requirements using apparent soil electrical conductivity (EC a ), digital soil map units, and relative elevation. Analysis indicated that the classifi cations explained <3% variability in yield response to N applied in the near-optimal range, where probabilities of receiving positive and negative marginal returns were the same. Presenting probabilities of yield response observed at diff erent ranges of N fertilization may provide the basis for assessing the uncertainty associated with the variable eff ects of weather and variable supply of N when assessing economic risk and benefi ts of N fertilization in large-scale on-farm studies. P.M. Kyveryga, On-Farm Network, Iowa Soybean Association, 4554 114th Street, Urbandale, IA 50322; A.M. Blackmer, in memory, Dep. of.