Tumors of the epithelial surface account for about 80% of all ovarian neoplasms and exhibit a heterogeneous histological classification affecting survival. Tumors of low malignant potential, defined as borderline ovarian tumors (BOTs), have a markedly better survival and low recurrence, even if surgery still represents the common management for this type of cancer. It is still debated in the literature if BOTs can be considered as intermediate precursors in the progression to high grade ovarian tumors. Evidences now propose that high-grade serous carcinomas are not associated with a defined precursor lesion. Together with histopathological studies, mutations of KRAS, BRAF and p53 genes, microsatellite instability (MSI) and under- or over-expression of many genes and proteins have been used to address this question. Despite the large body of data summarized, a limited number of molecules proved to be useful in elucidating BOTs pathogenesis and only a few of these showed possible application in the therapy. We believe that high-throughput technologies would help to overcome these limitations offering the promise of a better understanding of BOTs classification. The aim is to guide the diagnosis and prognosis of BOTs to develop new possible therapeutic molecular targets avoiding surgery.