The hydrology of the riparian zone plays an important role in regulating the water quality of agricultural watersheds. While slow flow pathways such as matrix flow offer opportunities for nutrient removal from water, faster flow pathways resulting from preferential flow often circumvent these treatment opportunities. Our objective was to evaluate matrix versus preferential flow in seep and non-seep areas of the riparian zone in a small mixed land use watershed in central Pennsylvania. Three seeps were selected in a 40-ha watershed. A field of suction cup lysimeters and piezometers were installed in each seep and in an adjacent non-seep area. Each field (10 m x 10 m) consisted of a grid of 20 piezometers and 12 suction cup lysimeters installed at both 20-cm and 60-cm depths and one piezometer installed below (to 4-m depth) the fragipan. Water samples from the piezometers and lysimeters have been collected three times since July 2010, with plans to collect samples during storm and non-storm periods four times annually. Spatial variability of water nitrate concentration was more than three times greater in the seeps compared to the non-seep areas; though mean nitrate-N concentration was relatively low between July and Nov. 2010 (0.34-1.05 mg L^-1 for the seeps; 0.11-0.25 mg L^-1 for non-seep areas). Total N was consistently greater in the seeps than in the non-seep areas (3.0 vs 2.1 mg L^-1). Chloride content was greater in the seeps than in the non-seep areas (5.3 vs 2.9 mg L^-1), as was spatial variability in Cl content (4.8 vs 1.5 mg L^-1). These preliminary results suggest that the seep water source (greater Cl) is unlike the water source of the adjacent non-seep areas. Greater spatial variability in nitrate and Cl concentrations in these seeps point to the importance of preferential flow paths.