Introduction <p>Early diagnosis of ovarian cancer, using cost-effective and non-invasive methods remains an unmet medical need, largely due to unspecific symptoms of the disease.</p> Objective <p>Our goal was to identify differentially methylated CpG loci between cervical swabs obtained from patients diagnosed with benign ovarian disease and with malignant pelvic mass.</p> Methodology <p>Using Infinium EPICv2 array, we interrogated methylation profiles of 77 cervical swabs. The study cohort was then divided into a training and testing set to develop a diagnostic signature. We applied several strategies to pinpoint CpG sites able to differentiate cervical swabs obtained from ovarian cancer patients and patients with benign ovarian disease.</p> Results and conclusions <p>None of the statistical methods applied identified CpG loci capable of diagnosing ovarian cancer with sufficient specificity and sensitivity. We conclude that methylation differences observed do not adequately distinguish between benign and malignant ovarian disease. The variations attributable to clinical conditions are likely obscured by the differences in cell composition, which is the primary source of sample heterogeneity. Therefore, we suggest that diagnostic tools should not rely on local methylation profile of the cervix but rather focus on detecting cancer-specific sequences transferred from the tumor site and present in cervical swabs.</p> Non-technical summary <p>Ovarian cancer is difficult to detect early, and we aimed to explore whether DNA methylation in cervical swabs could serve as a diagnostic tool. However, our study found that methylation patterns in these samples do not reliably distinguish between benign and malignant conditions, likely due to variations in cell composition. We recommend future research focus on detecting tumor-specific DNA sequences in cervical swabs instead.</p>