Riparian buffers are a best management practice (BMP) implemented to improve water quality. In 1997, Maryland established the Conservation Reserve Enhancement Program (CREP) to give landowners incentives to install riparian buffers that would help restore the Chesapeake Bay. Although many studies support riparian buffers as a BMP, many have also reported a wide range of nutrient reductions. It is uncertain what factors control buffer function, yet they continue to be installed with high expectations. Water quality predictions become less accurate in hydrogeologically complex systems such as the Ridge and Valley (R&V) physiographic province. The purpose of this research was to assess the riparian buffer’s nutrient removal function of dissolved nitrogen and phosphorus in the R&V to understand the hydrologic controls further. Throughout western Maryland, we conducted two synoptic stream chemistry studies that contained forest buffers planted under CREP and a range of pre-existing natural forested riparian zones. We used a steady-state reach mass balance model to estimate lateral groundwater inputs and tested several nutrient models to describe the nutrients in groundwater discharge. We then aimed to understand if incentives given through CREP to landowners were adequate by performing a benefit-cost analysis (BCA) using three scenarios. We used the BCA results to estimate nutrient reduction costs using results from the Chesapeake Bay Watershed Model (CBWM) and our synoptic studies. Streams along CREP sites did not show strong evidence of nutrient retention. However, those containing a mix of natural forests with planted buffers showed significant nutrient declines in both synoptic studies. Several models tested (i.e., The Nature Conservancy model, Gburek and Folmar (1999), our base model) inadequately described nutrient discharge; however, our actual flow model performed best. Our BCA results found newly planted forest buffers under CREP provide the greatest financial gains to landowners, but grass buffers are the most cost-effective practice based on CBWM’s estimated nutrient reductions. Although our research did not assess grass buffers, our synoptic studies showed little indication that newly planted forest buffers significantly reduce nutrients in the R&V, suggesting stream water quality greatly depends on the watershed’s hydrogeomorphology that controls how major contributing sources filter through the landscape.