A field experiment was conducted during the 1991 growing season to test the hypothesis that cotton lint yield is maximized within the limits of the water supply if a H₂O : N ratio of 0.25 kg N ha^-1mm^-1 is maintained. Four water supplies and four N supplies were established using a factorial design under a center pivot. Growth analyses and plant maps were used to define treatment effects on production of fruiting sites and retention of fruit. Within the constraints of the 1991 growing season, water supply was the only treatment variable to have a significant effect on cotton productivity. Increasing water supply increased the production of fruiting sites with the major effect on position 2 and 3 sites. Fruit retention was affected to some extent with the major effect also being on position 2 and 3 sites especially between nodal positions 10 to 15. No significant independent effect of N supply nor any significant water x nitrogen interaction existed for production or retention of fruit. Final lint yields were only 60 - 70% of expected reflecting the poor growing conditions in late August through September. We believe these conditions negated the effect of N supply on cotton productivity.