Viruses of microbes are major players in various ecosystems, typically affecting the structuration and dynamics of microbial communities. They nevertheless remain poorly explored in anaerobic digestion (AD) plants, where biowaste is valorised into methane-rich biogas. The AD process ecosystems are very sensitive to disturbance, leading to inhibition and loss of methane production. We were therefore interested in better understanding the interplay between abiotic disturbance, microbial community composition, including the viromes, and process performance. The variations of both viral and prokaryotic populations were followed in batch AD microcosms under abiotic stress. Four types of abiotic disturbances were tested in triplicates. During incubation, either NaCl, NH4Cl or phenol was injected into the reactors, as inhibitors previously reported in full-scale plants. Mitomycin C was also tested, since it can activate the lytic cycle of proviruses. We first confirmed a significant impact of the tested stresses on biogas production. We performed 16S rDNA metabarcoding targeting archaea and bacteria, and shotgun metagenomic sequencing of 30 selected metaviromes. Metavirome coassembly showed a N50 of 3,886 bases, and yielded more than 105 contigs longer than 1000 bp. Among them, 34,947 were predicted as putative viral contigs using VIBRANT (including 823 of high quality). There was a good agreement between the prokarotyic community composition and the predicted hosts of the viral contigs, with the dominance of Clostridiales, in both cases. We also detected auxiliary metabolic genes related to carbohydrate metabolism and the sulphur relay system, which seems relevant to this specific ecosystem. In the near future, we will identify the viral contigs that are differentially abundant according to the applied disturbance, and interpret the observed community dynamics. We also plan to apply the epicPCR technique to confirm specific virus-host pairs.