Large quantities of nitrate from the Minnesota River Basin are contributing to hypoxia in the Gulf of Mexico.  Nitrate in the Minnesota River originate in tile-drainage of agricultural soils. In these flat, high clay soils surface drainage is also common and contributes to sediment and phosphorus loading of the river. Generally, the adverse impacts of sediment and total P are on a local scale whereas the impacts of nitrate and soluble P are on a regional scale. This study assessed the trade offs between the presence or the absence of surface inlets on subsurface water quality for two different nutrient sources. The study was conducted on a set of drainage plots in a randomized split plot design at Lamberton, MN.  The soil at the experimental site was a Webster clay loam. The largest effect of eliminating surface drainage was an increase in subsurface flow.  Data from 2004 show an average annual tile flow of 7.9 cm yr^-1 in plots with both surface and subsurface drainage compared to 13.4 cm yr^-1 in plots with only subsurface drainage, a 70% increase. Corresponding nitrate-N loads were 6.2 kg ha^-1 yr^-1 and 9.9 kg ha^-1 yr^-1, a 60% increase. Differences in nitrate loading appear to be solely from increased subsurface water flow since flow weighted nitrate-N loading (0.78 kg ha^-1 cm^-1 yr^-1) from both drainage practices were identical.  Plots with manure had slightly higher nitrate-N losses (8.8 kg ha^-1 yr^-1) than plots fertilized with urea (7.3 kg ha^-1 yr^-1).  These results suggest that eliminating surface drainage would have a negative impact on nitrate-N loading of the river but would decrease sediment (800 and 1015 kg ha^-1 yr^-1), DMRP (0.2 and 0.4 kg ha^-1 yr^-1), and total phosphorus (0.3 and 0.6 kg ha^-1 yr^-1) loadings for the urea and the manure treatments, respectively.