Intercropping with legumes is known to increase the plant-available nitrogen (N) in soil, but can also increase leaching of NO3- to the groundwater. To minimise NO3- leaching risks, knowledge of the N-release processes is essential, including an estimate of the contribution of legumes to total NO3- concentrations in soil. Our objectives were to answer the questions: (1) whether the presence of legume roots increases N mineralisation, and (2) whether the proportion of legume-derived N in NO3- could be calculated with the help of natural abundance 15N in NO3-. We sampled soil monoliths of a Medicago x varia Martyn monoculture in August 2004 and set up three treatments: 'disturbance' (sieved to <2 mm), 'disturbance-roots' (sieved to <2 mm and visible roots removed), and 'control' (left untreated). During an incubation period of 70 days, an N-free nutrient solution was leached through the samples weekly. In the leachates we measured total N, total organic carbon, NO3-, and NH4+ concentrations. Six of the 13 sampling dates were chosen for N isotope analysis in NO3-. Nitrate was separated as AgNO3. During the incubation, 3 to 6% of the initial total mass of total N (192 to 274 mg N) in soil was mineralised. Nitrogen mineralisation followed zero-order kinetics independent of treatments. Mineralisation rates decreased in the order control (day 70: 3.7 microg NO3-N (mg Ninitial)-1)>disturbance-roots (2.6 microg NO3-N (mg Ninitial)-1)>disturbance (1.9 microg NO3-N (mg Ninitial)-1), indicating that mineralisation of legume roots did not play a major role in N mineralisation. The delta15N values jumped from ca. 3 per thousand to ca. 8 per thousand after 2 weeks of incubation, which we attributed to the contribution of two N pools. An exponential two-pool model could not be fitted to the data. Legume-derived soil organic matter, SOM (pool 1), was mineralised at the same rate as SOM accumulated before establishment of the legumes (pool 2). Fresh legume roots did not contribute significantly to N mineralisation.