In-stream sensors are increasingly deployed as part of ambient water quality-monitoring networks. Temporally dense data from these networks can be used to better understand the transport of constituents through streams, lakes or reservoirs. Data from existing, continuously recording in-stream flow and water quality monitoring stations were coupled with the two-dimensional hydrodynamic CE-QUAL-W2 model to assess the potential of altered reservoir outflow management to reduce sediment trapping in John Redmond Reservoir, located in east-central Kansas. Monitoring stations upstream and downstream from the reservoir were used to estimate 5.6 million metric tons of sediment transported to John Redmond Reservoir from 2007 through 2010, 88% of which was trapped within the reservoir. The two-dimensional model was used to estimate the residence time of 55 equal-volume releases from the reservoir; sediment trapping for these releases varied from 48% to 97%. Smaller trapping efficiencies were observed when the reservoir was maintained near the normal operating capacity (relative to higher flood pool levels) and when average residence times were relatively short. An idealized, alternative outflow management scenario was constructed, which minimized reservoir elevations and the length of time water was in the reservoir, while continuing to meet downstream flood control end points identified in the reservoir water control manual. The alternative scenario is projected to reduce sediment trapping in the reservoir by approximately 3%, preventing approximately 45 000 metric tons of sediment from being deposited within the reservoir annually. This article presents an approach to quantify the potential of reservoir management using existing in-stream data; actual management decisions need to consider the effects on other reservoir benefits, such as downstream flood control and aquatic life.