There exists a need (i) to test, whether equal or better cereal yields could be obtained using cropping systems which rely on renewable resources rather than on fertilizer nitrogen; and (ii) to discover the condition of the soil resource under these systems.The long-term cropping systems on a Gray Luvisol at Breton were studied. They included: (i) an agro-ecological 8-yr rotation (AER), established in 1981, which involved addition of both fababean green manure and manure from livestock fed with forages and fababeans grown in the rotation: (ii) a continuous grain (barley) system (CG), with fertilizer N at 90 kg ha−1 y−1, established in 1981; (iii) a classical Breton 5-yr rotation (CBR) involving forages and cereals, with no return of crop residues or manure, established in 1930. Mean barley yields were 16–19% higher in the AER (P ≤ 0.05) than in the CG system, and yield on either was about double that of the CBR. Within 9 yr, there was evidence of increased total C, N, and P; available N, P and K, CEC; microbial biomass, microbial respiration; and counts of bacteria, fungi, and mycorrhizae in the AER compared with the CG system.We conclude that biological fixation of N by legumes can be used as the sole source of N for barley production on Luvisolic soils of low fertility such as the Breton loam, without sacrificing yield or soil quality. Barley yields in the AER (38% of the rotation time) exceeded those of barley grown under continuous cereal cropping. The soil resource was maintained or improved during a 10-yr period under AER compared to the CG or CBR systems. Further research is needed to discover the mechanisms involved in regulating biological activity and availability of plant nutrients other than N in the AER system. Key words: Barley, Breton loam, cropping systems, Gray Luvisol, soil quality, fababeans