Among all fatal gynecological malignant tumors, ovarian cancer has the highest mortality rate. The purpose of this study was to develop a stable and personalized glycometabolism-related prognostic signature to predict the overall survival of ovarian cancer patients. The gene expression profiles and clinical information of ovarian cancer patients were derived from four public GEO datasets, which were divided into training and testing cohorts. Glycometabolism-related genes significantly associated with prognosis were selected. A risk score model was established and validated to evaluate its predictive value. We found 5 genes significantly related to prognosis and established a five-mRNA signature. The five-mRNA signature significantly divided patients into a low-risk group and a high-risk group in the training set and validation set. Survival analysis showed that high risk scores obtained by the model were significantly correlated with adverse survival outcomes and could be regarded as an independent predictor for patients with ovarian cancer. In addition, the five-mRNA signature can predict the overall survival of ovarian cancer patients in different subgroups. In summary, we successfully constructed a model that can predict the prognosis of patients with ovarian cancer, which provides new insights into postoperative treatment strategies, promotes individualized therapy, and provides potential new targets for immunotherapy.