Background: Glioma is the most frequent primary malignance of the central nervous system (CNS) and causes the top-ten deaths of cancers. Isocitrate dehydrogenase (IDH) mutations and chromosome 1p and 19q codeletion are the known prognostic factors of glioma. However, it is still not enough to predict the outcomes. Herein, we conduct a prognostic model of patients with gliomas. Methods: The Chinese Glioma Genome Atlas (CGGA) contained 286 patients with whole-exome sequencing (WES) was used to established the model. Kaplan-Meier analysis based on log-rank test was used to confirm the single mutation overall survival (OS) factors (allele frequency≥ 5) in CGGA. The model was established based on the Cox regression of above significant factors. The data of 1122 patients with glioma in TCGA (The Cancer Genome Atlas) was used to validate the model. Results: According to the CGGA database, the OS factors of glioma patients were 1p/19q, CIC, EGFR (p< 0.001), IDH1, IRS2, NF1, NOTCH1, PDGFRA, PTEN, ROS1, SMARCA4 and TP53. The Cox regression revealed 1p/19q codeletion (HR = 2.008, 95%CI 1.397-2.885, p< 0.001), EGFR (HR = 0.420, 95%CI 0.217-0.811, p= 0.010), PDGFRA (HR = 0.271, 95%CI 0.139-0.528, p< 0.001), and ROS1 (HR = 0.197, 95%CI 0.071-0.549, p= 0.002). Based on the β value of the regression equation, we divided the glioma patients into 3 groups (G1: harboring EGFR, PDGFR and/or ROS1 mutations; G2: wide type; G3: harboring 1p/19q without mutations in three genes). In TCGA, the OS of glioma patients was also showed G1 < G2 < G3 significantly. Conclusions: The prognostic model was divided glioma patients into three groups reasonably and could provide new reference for the glioma management. It needed to be further verified in more cohorts.