A total of 253 oral samples from different parts of China were collected prospectively. After a rigorous diagnosis and exclusion procedures, 235 samples completed the Miseq sequencing, 44 patients with CKD and 88 healthy controls samples from Zhengzhou, China, and 59 patients with CKD and 44 HCs samples from Hangzhou, China, were included. In the discovery phase, we characterized oral microbiome between 44 CKD and 88 HCs and identified the microbial markers and constructed a CKD classifier by a random forest classifier model between CKD and HCs. In the validation phase, 59 CKD and 44 HCs were used to validate diagnosis efficacy of CKD classifier and as independent diagnostic cohort to validate the diagnostic efficiency of CKD classifier; CKD, chronic kidney disease; HC, healthy control; RFC, random forest classifier model.