Subsurface drip irrigation was compared to furrow irrigation of cotton on 990-ft runs for 4 seasons on sandy soil. The first furrow irrigation after planting was difficult (as expected) and wasted a lot of water, up to 4 in. The drip system was less efficient than expected due to the startup time required. On the average for the season, the furrow system used 10.5 in. more water that the drip system, a difference which is very similar to results from published short-run (300 ft) tests. There was no difference in yield between the drip and furrow treatments. There were extreme variations in the infiltration rates across the 2 fields. An electrical conductivity (EC) cart, with rolling disk electrodes, was found to be a very fast and easy way to detect and precisely map areas of sharp changes in soil texture. Likewise, large differences in yield due to treatment, pests, or unusual soil conditions were quite obvious in our yield monitor work. Affected field areas were sharply delineated. The required weight coefficient, a correction factor for the yield monitor, was dependent on sensor blockage, and appeared to be a linear function of yield. Yields in a small calibration plot showed that severe moisture stress at peak bloom did not reduce yield, but the same stress one week after cut-out reduced yield drastically. By studying all of the available layers of spatially distributed parameters, such as soil properties, plant size and image reflectance, and yield, one can get a very good understanding of what's going on in a large field of cotton.