Epithelial ovarian cancer (EOC) is the most ominous tumor of gynecological cancers due to its poor early detection rate and unfavorable prognosis. To date, there is no reliable screening method for the diagnosis of ovarian cancer at an early stage. MiRNAs are small non-coding RNA molecules, and their main function is to regulate gene expression. The present study compared the serum miR-1181 and miR-4314 levels in patients with EOC and healthy controls to measure the diagnostic and prognostic value as candidate biomarkers. We collected serum samples from a total of 135 participants (69 patients with EOC and 66 healthy controls). Relative expressions of miR-1181 and miR-4314 were measured by quantitative real-time polymerase chain reaction assay (qPCR). The present study revealed that both serum miR-1181 and miR-4314 levels in patients with EOC were significantly increased compared to healthy controls for each marker. In addition, there was a significant relationship between miR-1181 and miR-4314 overexpressions and the stage and prognosis of the disease. Finally, patients with high expression levels of miR-1181 and miR-4314 had significantly shorter survival rates than those with low expression levels. The current study proposed that serum miR-1181 and miR-4314 could discriminate the EOC patients from healthy controls. In addition, both miR-1181 and miR-4314 may be predictive biomarkers for ovarian cancer prognosis. Further studies are needed to confirm the findings of the present study.