A three-year study was initiated in 2003 to investigate shallow subsurface nitrogen (N) and phosphorus (P) cycling in cropland and adjoining riparian buffer soils. Our objectives were to quantify N and P differences in cropland and buffers over the growing season, examine soil and landscape factors affecting N and P variation, and evaluate relative buffer effectiveness. Average soil water nitrate-N was consistently and significantly lower in buffers, with no significant differences among forest, grass, and grass-willow. With the exception of buffers located on glacial outwash, buffers had lower ground water nitrate-N relative to cropland. Ground water nitrate-N was lowest in forested buffers, whereas ammonium-N tended to be highest. Season average water table depth, dissolved oxygen concentration, and soil water nitrate-N at 50 cm explained 77% of season average cropland ground water nitrate-N variation. Dissolved oxygen alone explained 70% of buffer ground water nitrate-N variation. Large attenuation of ground water nitrate-N occurred where moderately to well drained corn fields transition to poorly and somewhat poorly drained riparian soils. Buffers appeared effective at reducing P movement, though some vertical and lateral P movement was observed. Season average soil water dissolved reactive P (DRP) at 25 cm was highly correlated with DRP at 50 cm for buffers and cropland. Morgan extractable P in buffers was significantly correlated with adjoining cropland P concentrations. Ground water DRP concentrations in corn fields commonly exceeded eutrophication thresholds, but average ground water total dissolved P and DRP were lower for buffers. Results suggest that buffers were effective at reducing soil water and ground water N and P movement at the landscape scale, and that variation in soil, hydrologic, and landscape characteristics affected buffer performance.