In Morocco, the use of seed legumes is limited because significant water deficits and the low availability of soil phosphorus (P) limit nitrogen fixation. However, little is known about the ability of faba bean-rhizobium symbiosis to fix nitrogen in P-deficient soils and to transfer fixed nitrogen (N) to intercropped wheat. Arbuscular mycorrhizal fungi (AM) and their extraradical hyphae networks play an important role in the facilitation process by promoting interconnectivity and the transfer of nutrients, such as N and P, between associated plants. The aim of this study was to analyse the impact of AM inoculation on N2 fixation and the transfer of fixed N from faba bean to intercropped wheat. Germinated faba bean and wheat seeds were transferred into 1–l pots filled with a P-deficient soil that was collected from the Haouz valley near Marrakech (Morocco). Plants from the two species were grown in pots in either pure or mixed stands under greenhouse conditions, and each cropping system was subjected to three mycorrhizal inoculation treatments with a non-inoculated (AM0) and two concentrations of Rhizophagus irregularis inoculants containing 1000 (AM1) or 2000 (AM2) spores pot-1. The 15N isotope dilution method was used to determine the amount and proportion of atmospheric N fixed by faba bean (Ndfa%) and the fixed N that was transferred to wheat. Mycorrhizal inoculation had a significantly positive effect on the shoot dry weights and total shoot N in faba bean, but not in wheat. The cropping system had no significant effect on the plant growth and total shoot N in both faba bean and wheat. The Ndfa percentage was very high in all of the treatments, varying from 86 to 91%. The total N fixed by faba bean was 27% significantly higher in the AM2 treatment compared with the AM1 and AM0 treatments for both cropping systems combined. The estimated proportions of fixed N that were transferred from faba bean to wheat were far higher in AM1 (50%) and AM2 (32%) treatments than