The dysbiosis of oral microbiota contributes to diseases such as periodontitis and certain cancers by triggering the host inflammatory response. Developing methods for the immediate and sensitive identification of oral microorganism is crucial for the rapid diagnosis and early interventions of associated diseases. Traditional methods for microbial detection primarily include the plate culturing, polymerase chain reaction and enzyme-linked immunosorbent assay, which are either time-consuming or laborious. Herein, we reported a persistent luminescence-encoded multiple-channel optical sensing array and achieved the rapid and accurate identification of oral-derived microorganisms. Our results demonstrate that electrostatic attractions and hydrophobic-hydrophobic interactions dominate the binding of the persistent luminescent nanoprobes to oral microorganisms and the microbial identification process can be finished within 30 min. Specifically, a total of 7 oral-derived microorganisms demonstrate their own response patterns and were differentiated by linear discriminant analysis (LDA) with the accuracy up to 100% both in the solution and artificial saliva samples. Moreover, the persistent luminescence encoded array sensor could also discern the microorganism mixtures with the accuracy up to 100%. The proposed persistent luminescence encoding sensor arrays in this work might offer new ideas for rapid and accurate oral-derived microorganism detection, and provide new ways for disease diagnosis associated with microbial metabolism.