Objectives: Several studies have demonstrated that alterations of gut microbiome composition are associated with production of gut-derived uremic toxins, which can lead to progression of kidney disease. In the prospective study, we investigated the association between the gut microbiome and diabetic kidney disease (DKD) progression. Methods: From 2017 to 2018, we enrolled adults aged 18–85 years with type 1 or 2 diabetic patients with non–dialysis-DKD. We conducted 16s rRNA sequencing on feces collected at baseline. During the follow-up of 36 months, DKD progression [defined as a reduction of estimated glomerular filtration rate (eGFR) over 15 mL/min /1.73 m2 or progression to end stage kidney disease] occurred in 40 out of 84 patients.  Results: There were no differences in age, blood pressure, hemoglobin a1c, duration of DM, body mass index between progression group and non-progression group. However, eGFR (35 vs 22 (progression vs non-progression; mL/min/1.73m2), P < 0.001) and urine protein-to-creatinine ratio (0.9 vs 3.7 (progression vs non-progression; g/gCr), P < 0.001) were different between two groups. We observed an enrichment in Bacteroidales in non-progression group, and at the species level, Clostridium bifermentans, were enriched in non-progression group. However, Clostridiales were markedly increased in progression group and at the species level, Alistipes indistinctus were markedly increased in progression group. These differences in microbial composition were associated with potential changes as inferred by predicting metagemones using PICUSt analysis. Specifically, MetaCyc pathways related to “pyruvate fermentation to acetate and lactate” and “pyruvate fermentation to isobutanol” were abundantenriched in progression group (Figure 1). Conclusions: Our findings have suggested that the individual gut microbiome strains influence subsequent kidney outcomes and may serve as the potential therapeutics in patients with DKD.