A dynamic optimization model which introduces an intertemporal nitrate-nitrogen residual function is used to derive and evaluate nitrogen fertilizer optimal decision rules for irrigated cotton production in the Southern High Plains of Texas. Results indicate that optimal nitrogen applications critically depend on initial nitrate-nitrogen levels and nitrogen-to-cotton price ratios. Also, the results indicate that single-year optimization leads to suboptimal nitrogen applications, which helps explain long-term cotton yield declines in the Southern High Plains of Texas; but single-year optimization does not significantly impact the net present value of returns of irrigated cotton operations.