Remotely sensed imagery can be used to calculate vegetative indices allowing managers to classify biomass and crop ‘vigor' and establish crop management zones. Site specific, variable rate application technology allows producers then to make precise applications of plant growth regulators in these management zones. Formulating site specific strategies for different management zones will require in-field crop monitoring to assess crop response to changes in inputs. In 2004 and 2005 we used the COTMAN system to ground-truth remotely sensed imagery of a highly variable commercial field in NE Arkansas. A replicated strip trial with 3 treatments was used to and evaluate crop response to plant growth regulator applications. Treatments consisted of 1) fixed (=fixed) rate, broadcast mepiquat chloride (MC) application, 2) variable rate MC application based on the acquired images, and 3) untreated check. Sampling areas were selected in crop management zone classifications based on the GNDVI classifications made from June 2004 and 2005 aerial images. High, Medium, and Low biomass classifications were considered indicators of crop vigor with the high biomass zone classification reflecting our estimate of greatest vigor. Three sites were selected to monitor crop growth in each management zone. Through the season, crop monitoring, plant sampling, and hand harvesting were made at these georeferenced sites; there were 81 sites total (3 sites/zone*3 zones/MC treatment*3 MC treatments*3 replications). Plant height, sympodial development and 1st position square and boll retention were monitored from early squaring through physiological cutout using the COTMAN Squaremap procedure. In 2005, tarnished plant bug population densities were monitored weekly as well using drop cloth sampling. Additional data collected during flowering included petiole sampling for NO3-N determinations. Final plant mapping was performed following defoliation using the COTMAP protocol (Bourland and Watson, 1990, Crop Sci. 39: 224-226). Ten plants in one row per plot were examined for node number of first (lowest) sympodial branch on the main axis, number of monopodia, and number of bolls on sympodia arising from monopodia. Harvest data were collected in a 20 ft section of row in each site. In two 5 ft sections, boll counts were made and plants hand harvested. Fifty boll samples also were collected from consecutive plants and fruiting sites for HVI analysis. Results from 2005 will be summarized and compared with the previously reported 2004 findings.