Excess precipitation in Midwest agricultural production areas is often removed artificially via subsurface drainage systems that intercept and divert it to surface waters. Nitrogen (N), either applied as fertilizer or manure or derived from soil organic matter, can be carried as nitrate with the excess water in quantities that may have deleterious effects downstream. A field study was initiated in 1989 in Pocahontas County, Iowa, on 0.05 ha plots of glacially derived clay loams. The objective of this three-phase study was to determine the effect of N application rate on NO3-N concentration and loss in a corn-soybean rotation over a wide range of weather conditions. Nitrogen-rate treatment phases with five seasons each (six for phase II) were imposed on subsurface-drained, continuous-flow-monitored plots over a 16-year period. Phase I N rates ranged from 0 to 168 kg N ha-1 in 56 kg N ha-1 increments. Separate plots were used for each crop in phase I, and significant NO3-N concentration differences were not observed between corn or soybean plots; this led to combining both crops in a split-plot configuration for phases II and III to study system effects. Phase II N rates ranged from 45 to 179 kg N ha-1 in 45 kg N ha-1 increments. Phase III was limited to two rates, 168 and 252 kg N ha-1. Average yearly flow-weighted NO3-N concentrations ranged from 3.9 mg L-1 (45 kg N ha-1, 1995) to 28.7 mg L-1 (252 kg N ha-1, 2001). Average flow-weighted NO3-N concentrations (in mg L-1) ranked by N rate were: 23.4 (252), 13.2 (179), 15.5 (168), 11.9 (134), 11.7 (112), 8.1 (90), 9.5 (56), 5.7 (45), and 8.9 (0). Losses were precipitation dependent and were reflective of individual seasons and rates imposed. Average flow-weighted NO3-N losses (kg ha-1) ranked by N rate and by phase were: 58 (168), 68 (112), 48 (56), 50 (no N) for phase I; 8 (179), 15 (134), 19 (90), 7 (45) for phase II; and 49 (252), 32 (168) for phase III. Results indicate that concentrations generally increased with rate; the effect on losses was variable due to disparity in drainage volumes among years. Corn yield during all periods showed a strong correlation between N rate and yield. As N rate increased, yield increased. It should be noted that at least 50% of the years showed limited yield response to N application above the next to the highest rates. To achieve average NO3-N concentrations less than 10 mg L-1 (USEPA drinking water standard) in subsurface drainage at this site, N application rates would need to be less than 112 kg N ha-1. Rates currently recommended for this area range from 112 to 168 kg N ha-1. Results from this study have significant implications for N fertilizer management and subsurface drainage NO3-N loss to surface waters in the state, the Mississippi River, and the Gulf of Mexico.