Stream restoration efforts in the United States are increasingly aimed towards water quality improvement, yet little process-based guidance exists to compare pollutant removals from different restoration techniques for variable site conditions. Excess nitrate (NO₃⁻) is a frequent pollutant of concern due to eutrophication in downstream water bodies such as the Chesapeake Bay. We used MIKE SHE to simulate hydraulics and NO₃⁻ removal in a 90 m restored reach of Stroubles Creek, a second-order stream in Blacksburg, Virginia. Site specific geomorphic, hydrologic, and hydraulic data were used to calibrate the model. We evaluated in-stream structures that induce hyporheic zone denitrification during base flow and in set floodplains that remove NO₃⁻during storm flows. We varied hydraulic conditions (winter base flow, summer base flow, storm flow), biogeochemical parameters (literature hyporheic zone denitrification rates and newly available inset floodplain removal rates) and boundary conditions (upstream NO₃⁻concentration), sediment conditions (hydraulic conductivity), and stream restoration design parameters(inset floodplain length). Our results indicate that NO₃⁻removal rates within the 90 m reach were minimal. Structure-induced hyporheic zone denitrification did not exceed 3.1% of mass flowing in from the upstream channel, was achieved only during favorable background groundwater hydraulic conditions (i.e. summer base flow), and was transport-limited such that non-trivial removal rates were achieved only when the stream bed hydraulic conductivity (K) was at least 10⁻⁴m/s. Inset floodplain nitrogen removal was limited by floodplain residence time and NO₃⁻ removal rate, and did not exceed 1% of in flowing mass. Summing these removals for both restoration practices over the course of the year based on the frequency of storm and summer base flow conditions yielded ∼2.1% annual removal. Achieving 30% NO₃⁻ removal required increasing the length of stream reach restored to 0.9 km–819 km (depending on hydraulic conductivity) and 3.8–46 km (depending on inset floodplain length and nitrogen removal rate)for in-stream structures during base flow and inset floodplains during storm flow, respectively. In oneof the first comparisons of process-based modeling to the Chesapeake Bay Program stream restoration guidance, we found that the guidance overestimated hyporheic NO₃⁻ removal for our modeled reach, but correctly estimated inset floodplain removal. Overall, our results indicate that in-stream structures and inset floodplains can improve water quality, but overall required level of effort may be high to achieve desired results.