Estimates of nitrate leaching derived from porous ceramic cups and soil analysis were compared in the winter of 1992/93 on farms within UK Pilot Nitrate Sensitive Areas. In these areas farmers have modified crop management with the aim of reducing nitrate losses to groundwater. These changes are relevant to ecological agriculture. The fields studied contained a range of arable crops, including cover and catch crops, and grass. Autumn soil mineral nitrogen to 90 cm (SMN) was highly correlated with estimates of leaching derived from porous cup samples, especially under arable crops. Correlations with the change in SMN over winter were less good and more affected by cropping situation. They were especially poor on grassland. These discrepancies were attributed to mineralisation of soil N and crop uptake during the leaching period. The effects of crop, soil and weather interactions on patterns of nitrate loss are discussed. The use of porous cups is preferred to SMN for estimating leaching since it avoids the need to allow for N transformations and is therefore more precise, especially when leaching losses are small.