The Swiss soil monitoring network (NABO) is based on more than 100 sites, which have been assessed every five year since 1985 using physico-chemical analyses. In order to expand soil quality assessment, analysis of soil microbial diversity was launched in 2012 on 30 of these sites. Standardization, reproducibility and cost-benefit-ratio of applied methods are crucial for long-term monitoring programs. Therefore, metabarcoding was evaluated from 2012 to 2016 by analysing yearly, triplicate samples from the 30 sites. Barcodes of the ribosomal operon were sequenced to determine soil microbial (bacterial and fungal) community structures (SMCS). Alpha-diversity revealed low correlations to environmental factors, with the highest value for pH (r = 0.51). Interactions between environmental and community structure data were more pronounced. Permutational analysis of variance showed significant differences between SMCS of different land-use types (arable land, grassland, forest), and between sites. Variability of SMCS was much lower over time than among sites. Time (sampling year) explained only 1% of data variance compared to more than 70% by the sites. These results showed that community data was better suited for monitoring purposes then sum-parameters like DNA-biomass or alpha-diversity. Furthermore, the stability of SMCS over several years allowed for the definition of site-specific baselines in the NABO system. Based on the numerous metadata collected at the sites, we intend to relate possible future changes of SMCS to environmental factors and to assign sensitive taxa to specific environmental stresses. The definition of such indicator taxa for soil quality is a long-term objective of our research.