Abstract Background and Aims Many uremic toxins (UTs) originate from gut microbiome, and contribute to chronic kidney disease (CKD) progression and cardiovascular morbidity. In order to reduce uremic symptoms and CKD progression, patients have several dietary restrictions, which may influence gut microbiome composition, and impact UTs production. An altered microbiome may contribute to UTs increase in those patients. However, the role of key bacterial taxa in producing UTs and the impact of diet on UTs variance in non-dialyzed patients are not well known. The objectives of this study were, first, to compare microbial features between CKD patients and healthy controls, and, second, to investigate the relation of gut microbiome with uremic toxicity, as well as the potential impact of diet on such relationship. Method Characterization of gut metagenomes, 10 UTs and 3 precursors’ serum concentrations by LC-MS/MS, host characteristics and diet were obtained from 240 non-dialysis CKD patients from the CKD-REIN cohort (mean ± SD): age: 68 ± 11 years, 71% male, estimated glomerular filtration rate (eGFR): 33.2 ± 12.7 ml/min/1.73m². First, to identify microbial biomarkers characterizing the gut microbiome-related toxicity in CKD, we compared microbiome features between 78 CKD patients and 78 age-, sex-, and BMI-matched healthy controls from the Milieu Interieur (MI) cohort: age: 58 ± 10 years, 60% male, eGFR: 89 ± 13. Second, we performed a multiomics’ data integration analysis via a supervised modelling to investigate cross-sectionally the association between host characteristics, gut microbiome, UTs, and diet-related features according to CKD severity (eGFR<30, n = 110 vs eGFR ≥30 mL/min/1.73m², n = 130). Results Compared to healthy controls, CKD patients had a significant reduced gut microbiome health index. Several Metagenomic Species Pan-genomes (MSPs) were significantly contrasted between MI and CKD cohorts: 43 species were enriched in CKD patients vs 24 in controls. Species most enriched in CKD patients included several UTs producers such as Lachnospiraceae spp, Dysosmobacter – Oscillibacter spp, Butyricimonas faecihominis, Victivallis vadensis and Hungatella spp, some of which were positively correlated with the following UTs: 3-Carboxy-4-methyl-5-propyl-2-furanpropionate (CMPF), trimethylamine-N-oxide (TMAO), and indole-3-acetic acid (3-IAA). Moreover, species belonging to Enterocloster and Hungatella genera (both members of Lachnospiraceae family) were found to be negatively correlated with eGFR. Among species associated with CKD severity, species carrying genes for UTs production were observed such as Desuflovibiro fairfieldensis, Bacteroides clarus and Blautia obeum along with increasing alcohol and hot drinks consumption, CRP and several UTs (kynurenic acid, indoxyl sulfate and Phenylacetylglutamine) levels. In contrast, some taxa like Faecalibacterium prausnitzii and Dysosmobacter welbionis were associated with legume intake but not with UTs. Conclusion Our study highlights an alteration of gut microbiome in CKD patients compared to healthy controls, with increased abundance of UTs producer species. The results of the multidimensional data integration modelling suggest a strong interplay between food intake, gut microbiome modifications, UTs accumulation and clinical features. These findings might open to promising therapeutic strategies to reduce microbiome-related toxicity.