We assess the predictive accuracy of Glover's (1974) stream-aquifer analytical solutions, which are commonly used in administering water rights, and evaluate the impact of the assumed idealizations on administrative and management decisions. To achieve these objectives, we evaluate the predictive capabilities of the Glover stream-aquifer depletion model against the MODFLOW numerical standard, which, unlike the analytical model, can handle increasing hydrogeologic complexity. We rank-order and quantify the relative importance of the various assumptions on which the analytical model is based, the three most important being: (1) streambed clogging as quantified by streambed-aquifer hydraulic conductivity contrast; (2) degree of stream partial penetration; and (3) aquifer heterogeneity. These three factors relate directly to the multidimensional nature of the aquifer flow conditions. From these considerations, future efforts to reduce the uncertainty in stream depletion-related administrative decisions should primarily address these three factors in characterizing the stream-aquifer process. We also investigate the impact of progressively coarser model grid size on numerically estimating stream leakage and conclude that grid size effects are relatively minor. Therefore, when modeling is required, coarser model grids could be used thus minimizing the input data requirements.