Aims: To identify a novel subtype with DNA driver methylation-transcriptomic multiomics and predict prognosis and therapy response in serous ovarian cancer (SOC). Methods: SOC cohorts with both mRNA and methylation were collected, and DNA driver methylation (DNAme) was identified with the MithSig method. A novel prognostic subtype was developed by integrating the information on DNAme and prognosis-regulated DNAme-associated mRNA by similarity network fusion. Results: 43 overlapped DNAme were identified in three independent cohorts. SOC patients were categorized into three distinct subtypes by integrated multiomics. There were differences in prognosis, tumor microenvironment and response to therapy among the subtypes. Conclusion: This study identified 43 DNAmes and proposes a novel subtype toward personalized chemotherapy and immunotherapy for SOC patients based on multiomics.