A number of fields on an organic (a potato field, a pasture and a field with red cabbage) and a conventional mixed farm (a potato field, a pasture and a field with cauliflower) were compared with respect to nitrate leaching in winter, using both soil mineral N measurements and a simulation approach. Soil organic matter N mineralization rates of the surface layers were determined during laboratory incubations. Mineral N contents of the 0-90 cm layer of each field were monitored during the period October-March. A coupled N mineralization-leaching model was used to simulate NO3- leaching over this period. Calculated NO3- leaching was in the same range on most fields, between 66 and 87 kg NO3--N ha(-1), except for the conventional pasture, which had smaller losses (35 kg NO3--N ha(-1)), and the conventional cauliflower, which had very large losses (293 kg NO3--N ha(-1)). The contribution of N mineralization during autumn and winter to NO3- leaching was important, and could only be taken into account explicitly using simulation models. The most remarkable difference with respect to N leaching was observed between the conventional cauliflower and organic red cabbage field, both with very high N fertilization rates. This was attributed to the differences in type of fertilizer applied, with N fertilization on the conventional fields much more conducive to N losses. The average NH4+-N concentrations during the field experiment were larger on the organic fields than on the conventional fields. This indication for a retardation or inhibition of nitrification in the organic fields is an important observation, which obviously could have important implications for NO3- leaching, but needs to be further investigated and quantified before it could be incorporated into simulation models.