Concentrations of NO - 2 -N in land drainage and river waters in Northern Ireland in recent years have frequently exceeded EEC guide values. Very little information exists to indicate if and when NO - 2 accumulates in soil solution, and whether NO - 2 from the soil profile is the source of NO - 2 in drainage and river waters. The occurrence of NO - 2 in the field was studied and laboratory incubation experiments carried out to determine the possible sources of NO - 2 in grassland soil. Field studies were carried out to determine the occurrence and spatial variability of NO - 2 in a grazed, grassland soil. Plots receiving either 100 or 500 kg N ha - 1 yr - 1 were systematically sampled in May and October 1992. Concentrations of NO - 2 in soil were highly variable and ranged from 0 to 2.747 μg N g - 1 , the data being significantly skewed to the right. Correlation matrices and stepwise multiple regression analyses showed relationships between NO - 2 and a number of soil variables. Nitrite appeared to be related to variables which indicated its occurrence as a result of nitrification of either fertilizer- or urine-derived NH + 4 . Nitrate was repeatedly correlated to NO - 2 concentrations, suggesting that both nitrification and nitrate reduction may be responsible for NO - 2 formation. Spatially, nitrite occurred at random, basic geostatics producing only one variogram, showing an increase in NO - 2 concentrations with an increase in distance between sampling points. There was no pattern to the distribution of NO - 2 with depth, indicating differences in the ratios of the rates of NO - 2 production and consumption. Numbers of NH 3 -oxidizers were consistently higher than numbers of NO - 2 -oxidizers, with some degree of variation between samples. The microbial aspects of NO - 2 formation are discussed, including partial recycling of NO - 2 via the NO - 2 pool, and possible causes of NO - 2 accumulation due to the inhibition of NO - 2 -oxidizing bacteria. Laboratory incubation studies were carried out in which measurable NO - 2 flushes were induced. Increasing soil pH and NH + 4 concentrations produced large NO - 2 flushes, which peaked after about 17 days of incubation, then rapidly declined. Soil incubated with urea produced NO - 2 -N concentrations equivalent to those encountered in the field, suggesting that NH + 4 oxidation accounts for a significant proportion of NO - 2 formed in this soil.