A study was conducted to determine the effect of conservation (notillage and ridge tillage) and conventional (moldboard plow) tillage systems on NO−3 loss through surface runoff and tile drainage. Nitrate concentrations and total volume of surface runoff and tile drainage from conventional tillage (CT), no-tillage (NT), and ridge tillage (RT) all planted in continuous corn (Zea mays L.), and Kentucky bluegrass (BG, Poa pratensis L.) treatments, were measured for 3 yr, 1989 to 1991. All corn tillage treatments received a total of 178.6 kg N ha−1 annually during the growing season. The volume of water drained through the tiles in the corn tillage systems always exceeded the volume in surface runoff, typically by factors of 2 to 4. Tile drainage was greatest from the CT treatments, least from BG, and approximately equal from RT and NT treatments in 1989 and 1990. Concentrations of NO−3 in tile water from CT, RT, and NT treatments exceeded the maximum recommended safe limit for drinking water (10 mg N L−1) in 79% of the leaching events, with flow-weighted concentrations between 12 and 17 mg N L−1 in 1989 and 1990. Flow-weighted NO−3 concentrations were only 1.2 and 2.6 mg N L−1 from BG in 1989 and 1990, respectively. The total NO−3 lost in tile water in 1989 was 18, 14, 14, and 1 kg N ha−1 from the CT, RT, NT, and BG treatments, respectively, whereas in 1990 there were 29, 20, 20, and 3 kg N ha−1 lost from the CT, RT, NT, and BG treatments, respectively. Nitrate losses in surface runoff were lower than in tile drainage, with maximums of 2.6 kg N ha−1 for the RT and NT treatments in 1989 and 5.5 kg N ha−1 for the RT treatment in 1990. In 1989 and 1990, both RT and NT treatments had greater yields and N uptake in grain than the CT treatment. A serious drought in 1991 limited corn yield, N uptake in grain, and NO−3 loss.