Nonpoint source contributions of nitrogen, particularly from agriculture, have become a serious concern for many watersheds in North Carolina. Recent regulatory action has increased the implementation of various best management practices (BMPs), particularly riparian buffer zones, for the purpose of reducing groundwater NO3-N pollution. However, the best design for such buffers has been the subject of great debate. The objectives of this project were to evaluate the relative effects of buffer vegetation and width on groundwater NO3-N removal and to determine if denitrification was the process most responsible. The main project consisted of four identically-designed buffer replications located on a farm in the Coastal Plain of North Carolina. The influence of vegetative type on buffer NO3-N concentration decreases were as follows; trees had an average decrease of 57% (from 8.79 to 3.78 mg NO3-N L-1), fescue had a decrease of 40% (from 6.33 to 3.77 mg NO3-N L-1), switchgrass had a decrease of 44% (from 5.52 to 3.09 mg NO3-N L-1), native vegetation had a decrease of 37% (from 6.47 to 4.07 mg NO3-N L-1), and the no-buffer control had a decrease of 27% (from 4.93 to 3.62 mg NO3-N L-1). These calculations are averages for each vegetation type from all of the wells from both widths and depths from all four buffer replications. For the 8 m buffer width, a total average NO3-N concentration decrease of 12% (from 9.97 to 8.75 mg NO3-N L-1) was observed for the intermediate well depth, while a 54% (from 5.26 to 2.41 mg NO3-N L-1) was observed for the deep well depth. For the 15 m buffer width, a total average NO3-N concentration decrease of 59% (from 6.42 to 2.61 mg NO3-N L-1) was observed for the intermediate well depth, while a 75% (from 4.31 to 1.06 mg NO3-N L-1) was observed for the deep well depth. Despite these apparent observed differences in the NO3-N concentration decreases, there were no overall statistically significant differences (p>0.05) between any of the veg.