Abstract: Objective:To construct a prognosis risk model model based on the competitive endogenous RNA(ceRNA)regulatory network of ovarian cancer to provide a reference for predicting the prognosis of ovarian can-cer patients and optimizing diagnosis and treatment plans.Methods:Ovarian cancer gene expression datasets were downloaded from the Gene Expression Omnibus database (GEO), “limma”software package was used to analyze and screen the differentially expressed genes of ovarian cancer, and the lncRNA-microRNA-mRNA(ceR-NA) regulatory network was constructed.Two datasets containing clinical prognosis information of patients were selected as the training set and the verification set to construct a prognosis risk model model of ovarian cancer.Firstly, univariate Cox regression analysis was utilized to screen out prognostic genes in the regulatory network, and secondly, the multivariate Cox regression analysis was used to construct prognosis risk model of ovarian can-cer.Furthermore, the survival curve (K-M curve) and receiver operating characteristic curve (ROC curve) were used to evaluate the effectiveness and robustness of the model.Finally, real-time fluorescence quantitative poly-merase chain reaction (RT-qPCR) was used to verify the expression level of the model genes in clinical samples.Results:733 differentially expressed genes in ovarian cancer were screened, and a ceRNA regulatory network containing 134 differentially expressed genes was suc-cessfully constructed by predicting the regulatory rela-tionships among the differentially expressed genes.A multi-gene prognostic model of ovarian cancer con-taining five key genes(WASF3, SNAI2, PDE8B, PCDH17, RNF128)was constructed by Cox regression.The KM curve showed that the overall survival rate of the high-risk group was better than that of the lower risk group(P< 0.01), the areas under the ROC curve in the 1st, 2nd and 3rd years of the training set were 0.702, 0.737 and 0.741, respectively, and the areas under the ROC curve in the 1st, 2nd and 3rd years of the validation set were 0.652, 0.653 and 0.667, respectively, confirming that the model had good predictive efficiency and robustness.Gene expression verification showed that WASF3, SNAI2 and PDE8B expressions were down-regulated in ovari-an cancer tissues and the differences were statistically significant(P< 0.05), and PCDH17 and RNF128 showed a down-regulation trend.Conclusion:The prognosis risk model constructed based on ceRNA regulatory network has good efficacy and robustness in the prognosis diagnosis of ovarian cancer, which is conducive to further guid-ing clinical treatment.