Preservation and sustainable use of soil biological communities represent major challenges in the current agroecological context. Indeed, most of soil ecosystem services results from biological functions particularly driven by taxonomic and functional assemblages of microbiological communities (i.e. nutrient cycling, soil aggregation, depollution, etc.). Consequently, soil microbial communities are logical candidates as effective indicators of soil quality and sustainability. But, good biological indicators must be associated with references that encompass an operating range of measured values that allow performing the desired diagnosis. Even if numerous studies have focused on soil microbial communities over the last two decades, due to the variety of ecosystems, sampling designs and methods, we still lack reference databases and diagnosis tools for the robust evaluation of agricultural practices impact on soil microbial communities To reach this goal, we used an extensive set of samples originated from the French Soil Quality Monitoring Network (RMQS). This monitoring network is based on a 16-km regular grid across the 550,000 km2 French territory representative of its pedoclimatic diversity. All sites have been geo-positioned with a precision <0.5m and the soil profile, site environment, climatic factors, vegetation and land use described. We determined: (i) the microbial biomass by quantifying the DNA extracted from soils, and (ii) the bacterial taxonomic richness (i.e. the number of Operational Taxonomic Units or OTUs). Both measures are good microbial indicators as they satisfy technical, practical and economic prerequisites (i.e. to be simple, rapid, reproducible, cheap, high- throughput, etc.). Based on this reference database, two innovative diagnosis tools were developed, using statistical predictive models according to environmental parameters (soil physico-chemical, climatic characteristics). Indeed, they provide reference values fitted for a given pedoclimatic condition, which is to be compared to the corresponding measured data for a robust diagnosis of soil biological quality. These tools were validated on external datasets and their applicability directly in the field was demonstrated on a farm network. In conclusion, these innovative tools providing reference values for a given pedoclimatic condition allow a robust diagnosis of soil biological quality.