In Denmark suction cup-measured nitrate concentrations in soil water and N leaching have been used as a means for monitoring the impact of changes in the consumption of commercial fertilizer and improved utilisation of nitrogen in manure by agriculture. The measurements are a part of the Danish Agricultural Monitoring Programme (NOVANA) and are carried out in five small agricultural dominated catchments covering differences in climate, soil types and farming practices within the country. Statistical analyses allowing for annual variations in climate showed that nitrate concentrations in root zone water from loamy and sandy catchments decreased by, respectively, 23 and 48% during 1991-2003, after which no further decrease occurred until 2010/11. The aim of this study is to describe how the year to year variations in percolation and field balance characterise N leaching for 30 soil water stations in the period 1991-2010. A general linear model for annual observations of N leaching, N leaching. = 3479. +. 0.140. ×. percolation. +. 0.080. ×. surface N balance. -. 1.737 year, explained 29% of the variation. When N leaching was calculated as a mean for each soil water station for the period 1991-2010, the field N balance explained about 44% of the variation in leaching between the 16 soil water stations located in loamy catchments. A lower proportion, 36%, was recorded for the 14 soil water stations located in sandy catchments, but the signature increased to 60% if an outlier field, having a large excess input of manure, was omitted. From these results for Danish catchments, we conclude that field N balances are able to predict the risks of leaching provided that measurements are integrated over 30 soil water stations or a relatively long monitoring period. However, the data showed that the leaching responses to the field balance are different for loamy and sandy soils and depend on farm type.