The nitrogen transfer between alfalfa and ryegrass was studied through isotope dilution at three different levels of N fertilization (20 mg N/pot, 200 mg N/pot, 400 mg N/pot) in a pot experiment using quartz sand as a substrate. An isogenic, nodulating, but non nitrogen fixing alfalfa line was used as a reference crop. Fixed N was transferred to the grass in the 20 mg N treatment and contributed markedly to the N nutrition of the grass (about 50% of the N in the plants). No transfer of fixed N could be detected in the higher fertilized treatments, although nitrogen fixation was only slightly inhibited by the presence of the fertilized mineral N. It is concluded that N transfer is strongly influenced by the N concentration in the substrate and transferred N contributes only slightly to the productivity of the legume/grass mixture under the given experimental conditions.