Surface water and groundwater are intimately connected by a two-way flux between the stream and the underlying aquifers; a complicated yet crucial relationship to represent in models. However, the National Water Model currently only considers a one-way flux, where groundwater can enter a stream but cannot return to the aquifer. A case study of the Northern High Plains Aquifer, USA is used to investigate the consequences of omitting two-way stream-aquifer fluxes on streamflow prediction capabilities of the National Water Model. Evaluation is based on the average error in streamflow between modeled values from the National Water Model and observed values from USGS gages comparing between reaches identified as losing or gaining for hydrologic extreme events, floods and droughts. For floods, the lack of a losing stream mechanism resulted in modeled flood response characterized by an earlier peak discharge and an overestimate of the observed flood volume for losing streams representing statically significant results compared to gaining streams. For droughts, there was no statistical difference between losing and gaining streams existed, however, the National Water Model did statistically significantly overestimate the streamflow compared to the observed USGS hydrograph. The systematic overestimate of streamflow by the National Water Model could be in part due to the lack of a losing stream mechanism which was on average 0.1% streamflow loss per river mile.