Animals and their associated microbiota share long evolutionary histories. However, it is not always clear how host genotype and microbiota interact to affect phenotype. We applied a hologenomic approach to explore how host-microbiota interactions shape lifetime growth and parasite infection in farmed Atlantic salmon (Salmo salar). Multi-omics datasets were generated from the guts of 460 salmon, 82% of which were naturally infected with an intestinal cestode. A single Mycoplasma bacterial strain, MAG01, dominated the gut metagenome of large, non-parasitised fish, consistent with previous studies showing high levels of Mycoplasma in the gut microbiota of healthy salmon. While small and/or parasitised salmon also had high abundance of MAG01, we observed increased alpha diversity in these individuals, driven by increased frequency of low-abundance Vibrionaceae and other Mycoplasma species that carried known virulence genes. Colonisation by one of these cestode-associated Mycoplasma strains was associated with host individual genomic variation in long noncoding RNAs. Integrating the multi-omic datasets revealed coordinated changes in the salmon gut mRNA transcriptome and metabolome that correlated with shifts in the microbiota of smaller, parasitised fish. Our results suggest that the gut microbiota of small and/or parasitised fish is in a state of dysbiosis that partly depends on the host genotype, highlighting the value of using a hologenomic approach to incorporate the microbiota into the study of host-parasite dynamics. ENA project includes all DNA and RNA raw fastq files, and bacterial MAG fasta files. Linked raw metabolomics data are available at MetaboLights (accession MTBLS8575): https://www.ebi.ac.uk/metabolights/editor/MTBLS8575/descriptors Analysis scripts are available at the project GitHub: https://github.com/jcbrealey/Brealey_etal_salmon_multiomics Article published at mSystems (open-access): https://doi.org/10.1128/msystems.01043-23