Objective: To correlate the genome-wide methylation signature of microRNA genes with dysregulated expression of selected candidate microRNA in tissue and serum samples of epithelial ovarian cancer (EOC) and control using quantitative reverse transcription polymerase chain reaction (qRT-PCR), and evaluation of EOC predictive value of candidate microRNA at an early stage. Methods: We performed Methylated DNA Immunoprecipitation coupled with NGS (MeDIP-NGS) sequencing of 6 EOC and 2 normal tissue samples of the ovary. Expression of selected microRNA from tissue (EOC=85, normal=30) and serum (EOC=50, normal=15) samples was evaluated using qRT-PCR. We conducted bioinformatics analysis to identify the candidate miRNA’s potential target and functional role. Results: MeDIP-NGS sequencing revealed hypermethylation of several microRNAs gene promoters. Three candidate microRNAs were selected (microRNA-34a, let-7f, and microRNA-31) from MeDIP-NGS data analysis based on log2FC and P-value. The relative expression level of microRNA-34a, let-7f, and microRNA-31 was found to be significantly reduced in early-stage EOC tissues and serum samples (p<0.0001). The receiver operating characteristic analysis of microRNA-34a, let-7f and miR-31 showed improved diagnostic value with area under curve(AUC) of 92.0 (p<0.0001), 87.9 (p<0.0001), and 85.6 (p<0.0001) and AUC of 82.7 (p<0.0001), 82.0 (p<0.0001), and 81.0 (p<0.0001) in stage III-IV and stage I-II EOC serum samples respectively. The integrated diagnostic performance of microRNA panel (microRNA-34a+let-7f+microRNA-31) in late-stage and early-stage serum samples was 95.5 and 96.9 respectively. Conclusion Our data correlated hypermethylation-associated downregulation of microRNA in EOC. In addition, a combined microRNA panel from serum could predict the risk of EOC with greater AUC, sensitivity, and specificity.