SUMMARY The effects of an intercrop catch crop (Italian ryegrass) on (i) the amounts and concentrations of nitrate leached during the autumn and winter intercrop period, and (ii) the following crop, were examined in a lysimeter experiment and compared with that from a bare fallow treatment. The catch crop was grown in a winter wheat/maize rotation, after harvest of the wheat, and incorporated into the soil before sowing the maize. A calcium and potassium nitrate fertilizer labelled with 15 N (200 kg N ha −1 ; 9.35 atom per cent excess) was applied to the winter wheat in spring. Total N uptake by the winter wheat was 154 kg ha −1 and the recovery of fertilizer‐derived N (labelled with 15 N) was 60%. The catch crop (grown without further addition of N) yielded 3.8t ha −1 herbage dry matter, containing 43 kg N ha −1 , of which 4.1 % was derived from the 15 N‐labelled fertilizer. Two‐hundred kg unlabelled N ha −1 was applied to the maize crop. During the intercrop period the nitrate concentration in water draining from the bare fallow lysimeters reached 68 mg N1 −1 , with an average of 40 mg N1 −1 . With the catch crop, it declined rapidly, from 41 mg N I −1 to 0.25 mg N I −1 , at the end of ryegrass growth. Over this period, 110 kg N ha −1 was leached under bare fallow, compared with 40 kg N ha −1 under the catch crop. 15 N‐labelled nitrate was detected in the first drainage water collected in autumn, 5 months after the spring application. The quantity of fertilizer‐N that was leached during this winter period was greater under bare fallow (18.7% of applied N) than when a catch crop was grown (7.1 %). In both treatments, labelled fertilizer‐N contributed about 34% of the total N lost during this period. With the ryegrass catch crop incorporated at the time of seedbed preparation in spring, the subsequent maize grain‐yield was lowered by an average of 13%. Total N‐uptake by the maize sown following bare fallow was 224 kg N ha −1 , compared with 180 kg ha −1 with prior incorporation of ryegrass; the corresponding values for uptake of residual labelled N were 3% (bare fallow) and 2% (ryegrass) of the initial application. Following the maize harvest, where ryegrass was incorporated, 22.7% of the previous year's labelled fertilizer addition was present in an organic form on the top 30 cm of lysimeter soil. This compares with 15.7% for the bare fallow intercropping treatment. Tracer analyses showed overall recoveries of labelled N of 91.7% for the winter wheat/ ryegrass/maize rotation and 97% for the winter wheat/bare fallow/maize rotation. The study clearly demonstrated the ecological importance of a catch crop in reducing N‐leaching as well as its efficient use of fertilizer in the plant‐soil system from this particular rotation. However, the fate of the organic N in the ploughed‐down catch crop is uncertain and problems were encountered in establishing the next crop of maize.