Academic research models for Chesapeake Bay have, traditionally, been forced with USGS inputs, flows and nutrient loads from 10 major rivers. These tributaries fail to account for 100% of the inputs entering the Bay. In contrast, models used for determining Total Maximum Daily Load for Chesapeake Bay are forced with output from a watershed model at thousands of locations, presumably, accounting for all these inputs. Our aim is to increase understanding of the impacts different forcing schemes have on water quality model simulation. Simulations were completed using three forcing approaches: 1) using “traditional” USGS-derived input from 10 major rivers; 2) using “concentrated” input from 10 major rivers derived from watershed model output; and 3) using “diffuse” input from 1117 rivers derived from watershed model output. Comparisons of these schemes revealed large impacts on simulations in Chesapeake Bay during periods of high flow and extreme weather events under diffuse forcing.