In response to the need for improved agricultural best management practices in irrigated lands, the GLEAMS(Groundwater Loading Effects of Agricultural Management Systems) model was modified to include a component that modelsfurrow, basin, and border irrigation practices. The irrigation component of the new model, GLEAMSIR, was validated withresults from SRFR, a full hydrodynamic irrigation model. Sensitivity of nutrient and irrigation output parameters to modelinput parameters was assessed with a singlevariable method and a stochastic method, Monte Carlo. Means and distributionsof soil parameters for a surfaceirrigated cotton field in Marana, Arizona, were used for sensitivity analysis. For bothsinglevariable and Monte Carlo sensitivity analyses, output parameters were most sensitive to infiltrated depth after120 min, permanent wilting point, and field capacity. A study of sludge application on the same field (1x and 3x sludgeapplication, nitrogen fertilizer application, and control with no nitrogen) was used to evaluate the GLEAMSIR model resultsfor different management scenarios and to compare GLEAMSIR results with nutrient concentrations at different locationsalong the furrow. The variability of measured nitrate concentrations along the furrow indicated the need to account forvariable infiltration along the furrow in GLEAMSIR.