Previous work has shown a strong correlation between the plant population of winter wheat and the coefficient of variation (CV) from sensor readings taken mid-season. It was also observed that in areas where plant stands were poor the addition of nitrogen was not beneficial. This study evaluated the use of CV to better predict the yield response to nitrogen using a response index (RI). The experiment was established at 3 locations. Treatment structure included 3 seeding rates (40, 80, and 120 kg ha-1) with 4 N rates (0, 40, 80, and 120 kg N ha-1). Experimental design was a RCB with 3 replications. Each plot was sensed at Feekes 3, 5, and 7 using the Green SeekerŪ handheld sensor. The correlation between predicted RI (determined by the normalized difference vegetative index) and RI yield was improved through the introduction of CV into the calculation of predicted RI. The application of mid-season N can be better refined through the implementation of the methods developed.