A novel saliva-based miRNA profile has been developed to diagnose and predict oral cancer (OC). The study utilized TCGA miRNA sequencing data and small RNA sequencing data from saliva samples to identify differentially expressed miRNAs. Eight miRNAs (miR-7-5p, miR-10b-5p, miR-182-5p, miR-215-5p, miR-431-5p, miR-486-3p, miR-3614-5p, and miR-4707-3p) were identified and validated in saliva samples from OC (n=50), oral potentially malignant disorders (OPMD) (n=52), and controls (n=60). The eight-miRNA signature demonstrated high diagnostic accuracy, with an area under the curve (AUC) of 0.954, sensitivity of 86%, specificity of 90%, positive predictive value (PPV) of 87.8%, and negative predictive value (NPV) of 88.5% for distinguishing OC from controls. For OC vs. OPMD, the AUC was 0.911, sensitivity 90%, specificity 82.7%, PPV 74.2%, and NPV 89.6%. A risk probability score was developed to predict OC risk in OPMD patients. The study also identified a four-miRNA panel (miR-10b-5p, miR-182-5p, miR-215-5p, miR-3614-5p, and miR-4707-3p) that effectively discriminated OPMD from OC. The four-miRNA panel achieved an AUC of 0.9115, sensitivity of 90%, specificity of 82.7%, PPV of 74.2%, and NPV of 89.6%. The study highlights the potential of saliva-based miRNAs as non-invasive biomarkers for OC diagnosis and risk prediction, offering a promising approach for early detection and management of OPMD. The results suggest that saliva-based miRNA signatures could revolutionize the clinical management of OC by enabling early diagnosis and risk stratification.A novel saliva-based miRNA profile has been developed to diagnose and predict oral cancer (OC). The study utilized TCGA miRNA sequencing data and small RNA sequencing data from saliva samples to identify differentially expressed miRNAs. Eight miRNAs (miR-7-5p, miR-10b-5p, miR-182-5p, miR-215-5p, miR-431-5p, miR-486-3p, miR-3614-5p, and miR-4707-3p) were identified and validated in saliva samples from OC (n=50), oral potentially malignant disorders (OPMD) (n=52), and controls (n=60). The eight-miRNA signature demonstrated high diagnostic accuracy, with an area under the curve (AUC) of 0.954, sensitivity of 86%, specificity of 90%, positive predictive value (PPV) of 87.8%, and negative predictive value (NPV) of 88.5% for distinguishing OC from controls. For OC vs. OPMD, the AUC was 0.911, sensitivity 90%, specificity 82.7%, PPV 74.2%, and NPV 89.6%. A risk probability score was developed to predict OC risk in OPMD patients. The study also identified a four-miRNA panel (miR-10b-5p, miR-182-5p, miR-215-5p, miR-3614-5p, and miR-4707-3p) that effectively discriminated OPMD from OC. The four-miRNA panel achieved an AUC of 0.9115, sensitivity of 90%, specificity of 82.7%, PPV of 74.2%, and NPV of 89.6%. The study highlights the potential of saliva-based miRNAs as non-invasive biomarkers for OC diagnosis and risk prediction, offering a promising approach for early detection and management of OPMD. The results suggest that saliva-based miRNA signatures could revolutionize the clinical management of OC by enabling early diagnosis and risk stratification.