Hydraulic conductivity of the soil matrix dynamically responds to changes in the surrounding environment. Therefore, infiltration parameters for the Green-Ampt equation should change for each storm event in continuous simulation models. This study focused on improving Water Erosion Prediction Project (WEPP) model estimates of runoff using over 220 plot-years of natural runoff plot data from 11 locations. By optimizing the effective Green-Ampt hydraulic conductivity, K e , for each event within the simulation, a method of correlating hydraulic conductivity on any given day to many other parameters was established. Factors with significant correlation to optimized values of K e fell into three distinct categories; (1) factors related to soil crusting and tillage; (2) factors related to event size; (3) factors related to antecedent moisture conditions. Equations were developed to represent the temporal variability of hydraulic conductivity for each group. The equation describing the decrease in hydraulic conductivity owing to crusting used an exponential decay function based primarily on cumulative rainfall kinetic energy since last tillage, a soil stability factor, and a crust factor. The relationship between hydraulic conductivity and event size was characterized using an exponential relationship with total rainfall kinetic energy. The final adjustment used the moisture content immediately below the infiltration zone to account for the influence of antecedent moisture conditions on optimized hydraulic conductivities. All three adjustments were incrementally incorporated into WEPP and each improved the average model efficiency.