The growth regulator Pix is often used to reduce vegetative growth and enhance cotton's fruit production. Pix was applied to cotton (Gossypium hirsutum L.) in multiple applications totaling eight ounces/A in increments of two, four and eight ounces. Those treatments were compared to an untreated control. The test was conducted at the Tennessee Valley Substation in northern Alabama. Yields and defoliation dates were compared to those predicted by the Gossym-Comax cotton computer model. The two and four ounce Pix treatments were initiated at cotton's early square and repeated at one and two week intervals respectively. The eight ounce treatment was applied at early bloom. Two nitrogen (N) fertilizer rates were also used: 70 lbs N/A preplant and 70 lbs N/A preplant with 40 lbs N/A sidedressed. At early boll opening, ten plants from each plot were mapped for fruiting sites and boll set. Although all Pix applications increased cotton yields compared to the control, the only statistically significant increase (100 lbs lint/A) was found at 110 lbs N/A and four ounce Pix rate. The plant mapping data could not pinpoint the reason for this yield response, however, all Pix treatments did increase boll retention on fruiting branches six through ten. Surprisingly, plant mapping data of Pix treatments indicated that the number of fruiting sites and boll retentions were higher on fruiting branches 11 through 15 with 70 lbs N/A compared to 110 lbs N/A. Apparently the dry August weather condition caused these upper bolls to have little effect on yield. Using 110 lbs N/A Gossym-Comax accurately predicted cotton yields (90% of actual) and defoliation data for all Pix treatments. With 70 lbs N/A, however, Gossym-Comax's predicted yields were only 70% of actual yields. Gossym Comax's predicted defoliation data was off by as much as 8 days.