454-Pyrosequencing and lipid fingerprinting were used to link anaerobic digestion (AD) process parameters (pH, alkalinity, volatile fatty acids (VFAs), biogas production and methane content) with the reactor microbial community structure and composition. AD microbial communities underwent stress conditions after changes in organic loading rate and digestion substrates. 454-Pyrosequencing analysis showed that, irrespectively of the substrate digested, methane content and pH were always significantly, and positively, correlated with community evenness. In AD, microbial communities with more even distributions of diversity are able to use parallel metabolic pathways and have greater functional stability; hence, they are capable of adapting and responding to disturbances. In all reactors, a decrease in methane content to <30% was always correlated with a 50% increase of Firmicutes sequences (particularly in operational taxonomic units (OTUs) related to Ruminococcaceae and Veillonellaceae). Whereas digesters producing higher methane content (above 60%), contained a high number of sequences related to Synergistetes and unidentified bacterial OTUs. Finally, lipid fingerprinting demonstrated that, under stress, the decrease in archaeal biomass was higher than the bacterial one, and that archaeal Phospholipid etherlipids (PLEL) levels were correlated to reactor performances. These results demonstrate that, across a number of parameters such as lipids, alpha and beta diversity, and OTUs, knowledge of the microbial community structure can be used to predict, monitor, or optimise AD performance.