DNA methylation markers in the peripheral blood are able to be applied to treat epithelial cancer. Nevertheless, the diagnostic potential value of it for ovarian cancer (OV) has not been studied. The study aimed to explore the difference of DNA methylation in peripheral blood between OV patients and healthy women. Firstly, the whole blood of DNA methylation data was provided by the Gene Expression Omnibus (GEO) database. The linear model was applied to the identification of significantly differentially expressed methylated CpG sites (differentially methylation sites [DMP]), and the further screen of co‐expression CpG sites (Co‐DMP). A total of 2812 DMPs were identified, and weighted gene co‐expression network analysis helped to obtain seven co‐expression modules. Among them, the yellow module was the most related to OV. Co‐DMPs (167) in the yellow module were mainly distributed near the transcription start sites. However, most of them were not in the CpG island. Least absolute shrinkage and selection operator (LASSO) regression analysis was applied to the identification of stable OV‐related blood biomarkers that six Co‐DMPs (cg00134539, cg00226923, cg25268718, cg25697314, cg25839227, cg26574610) with the highest frequency were found as potential biomarkers. Finally, the diagnostic classifier was established using the support vector machine (SVM) with the accuracy rate of 87.1% and 74.5% in training data set and validation data set, respectively. To sum up, a new feature was provided here for the diagnosis of OV, which is helpful for the diagnosis and individualized treatments of early OV.