Inflammatory bowel disease (IBD) is associated with an inappropriate immune response to the gut microbiota and disruption of intestinal homeostasis. IBD patients and experimental animal models have consistently shown alterations in the gut microbiota composition. However, these studies have mainly focused on faecal microbiota samples taken after the onset of inflammation and IBD establishment. The colonic microbiota inhabits both the gut lumen and the mucus layer covering the intestinal epithelium. Thus, information about mucus-resident microbiota is not necessarily conveyed in the routine microbiota analyses of faecal samples. To address potential changes in microbial composition and function before the onset of IBD, we compared both mucus and faecal microbiota in the mdr1a-/- spontaneous model of colitis over times that we histologically defined as before onset of colitis, during and after colitis onset. We showed that alterations in microbiota composition preceded the onset of intestinal inflammation and that these changes were evident in the mucus, but not in faeces. This altered microbiota composition was coupled with a reduced inner mucus layer, indicating a compromised mucus barrier prior to colitis development. Upon emergence of inflammation, compositional differences were found in both mucus and faecal microbial communities. Spatial segregation of microbiota with intestinal mucosa was also disrupted on disease onset which we hypothesise contributes to a more severe intestinal pathology. Therefore, our data indicate that microbial changes start locally in the mucus and then proceed to the faecal matter concomitantly with colitis development.Next, we examined whether microbial gene functional potential and endogenous metabolite profiles followed alterations in gut microbiota taxonomic composition. Our findings showed that the microbial gene content was similar between mdr1a-/- mice and wild-type littermate controls, demonstrating stability of the gut microbiome at the face of ensuing gut inflammation. In further support of these findings, urinary metabolite analysis revealed that metabolite profiles were unaffected by intestinal inflammation. Metabolites previously reported to change in IBD were similar between mdr1a-/- and wild-type mice at stages preceding and during inflammation. We also found that changes in metabolite profiles did not correlate with colitis scores. However, metabolite changes could discriminate mdr1a-/- mice from wild-type controls, suggesting they could have value in predicting risk of IBD with a potential clinical use in at least a subset of individuals with MDR1A polymorphisms.To assess whether changes in antimicrobial proteins (AMPs) accounted for observed differences in mucus microbiota composition, we also investigated the expression of regenerating islet-derived protein 3 γ (Reg3γ), angiogenin 4 (Ang4), β-defensin 1 and resistin-like molecule beta (Relm-β) in the colon. We found similar levels of these AMPs as well as IgA-producing plasma cells between mdr1a-/- and wild-type mice, suggesting that other factors contribute to alterations in microbiota composition.Overall, our data indicate that the mdr1a-/- is a good model of colitis, as it enables us to look at pre-clinical changes in the gut microbiota. This work suggests the importance of mucus sampling for sensitive detection of microbiota changes. Furthermore, metabolite profiling may be a helpful way to discriminate genetic susceptibility to disease.