Understanding the spatial variability of cotton fiber quality at the field level is of paramount importance in order to deploy precision agriculture to fiber quality management in cotton production. The objective of this two-year study is to document the spatial variability of cotton fiber quality in a dry-land field near College Station, TX. In 2005, a 50-m regular grid containing 36 sampling nodes was established in the field. In 2006, an additional set of 30 nodes was added to the grid, yielding a total of 66 sampling points. In both years, cotton samples were hand-picked from each sampling point, ginned, and subject to the HVI classing to obtain various fiber quality properties such as micronaire, length, uniformity, etc. Semivariance analysis revealed that strong spatial dependence existed in most of the HVI fiber quality properties. Based on this, point kriging was performed to generate high resolution fiber quality maps. These fiber quality maps are suggestive of high- versus low-fiber-quality-potential areas in the field. They would also provide useful information to manage fiber quality site-specifically and plan harvest routes for better fiber segregation. Given the important role that bale-level fiber quality plays in cotton pricing and trading, farmer's monetary return would be increased.