Precision agriculture technologies offer an opportunity to vary production inputs within a field. Variable rate application offers the potential to increase production efficiency and minimize potential adverse environmental effects of agricultural chemicals. As an initial step in the development of precision agriculture technologies for cotton, studies are needed to document variability of cotton. The primary objective of this study was to document variability of yield and quality of irrigated cotton within and across three growing seasons. This study was conducted on a 5.3 ha irrigated field located at the Erskine Research Farm at Texas Tech University, Lubbock, TX. The crop was grown under a conventional tillage system with a 1.0 m row spacing. With the exception of sample collection, the field was managed traditionally with respect to production inputs. A grid system (57 points) was established on 30.5 m (approximately 0.1 ha) intervals. Production of fruiting sites, fruit retention, lint yield, fiber length, strength, micronaire, and gross revenue were estimated for each grid point. Soil chemical and physical properties were also determined for each grid point. Highest variability was observed for lint yield and production of fruiting sites, and lowest variability was observed for lint quality parameters. Yield was correlated to production of fruiting sites each season. Nitrate concentrations were highly variable, and yield was negatively correlated to nitrogen (N) in 1997. This suggests that variable application of N may be a viable management option in the future. Yield was positively correlated to calcium (Ca), pH, and CEC in 1997. Yield variability was correlated across growing seasons. Gross revenues were quite variable, due primarily to yield variability.