In recent years, local governments have invested heavily in management practices to reduce nutrient and sediment loads. These practices provide localities with nutrient and sediment regulatory reduction credits; however, their effects on water quality are poorly understood at the watershed scale. Long-term watershed-scale monitoring is essential for assessing progress toward water-quality goals, yet it has historically been lacking in urban watersheds. Since 2007, Fairfax County, Virginia, has partnered with the US Geological Survey to monitor and evaluate water-quality conditions in 20 small urban streams. This study assessed nutrient and suspended sediment loads, trends in concentration, and trends in load. Trends in load are affected by streamflow-induced variability that must be removed through a process called “flow-normalization;” however, existing methods have neither been applied to small urban watersheds nor to loads computed on a sub-daily timestep. In this study, four such methods also were assessed, and an adaptation of the weighted regressions on time, discharge, and season approach was found to be most effective. Loads, concentrations, and trends in load were spatially and temporally variable. Differences were attributed to physical watershed features such as geology, soils, and channel geomorphology, as well as urban sources such as turfgrass fertilization and septic infrastructure. Most notably, flow-normalized suspended sediment, nitrogen, and phosphorus loads decreased in two watersheds with completed stream restorations and increased in those with few implemented practices.