High-grade serous carcinoma (HGSC) is the most widespread and aggressive histotype of ovarian cancer. In recent decades, significant efforts have been made to improve HGSC survival rate and develop personalized therapy. Nevertheless, HGSC-related mortality remains high; therefore, identifying effective predictive biomarkers and target agents is essential. Four transcriptomic HGSC subtypes (mesenchymal, immunoreactive, differentiated, and proliferative) have been clustered using molecular and genetic methods, and an independent platform has been developed for their classification. These subtypes have individual genetic characteristics and demonstrate different clinical courses and prognoses; the mesenchymal subtype showed the poorest survival rate, whereas the immunoreactive subtype showed the best survival rate. Additionally, different molecular subtypes have different sensitivities to therapeutic agents. Thus, patients with the mesenchymal subtype can be treated more effectively using taxans, while patients with the immunoreactive subtype benefit more from immunotherapy. Because complicated genetic tests have low availability for routine practice, histology and immunohistochemistry-based classifications have been proposed for HGSC molecular subtyping. Moreover, deep learning-based diagnostic and prognostic algorithms have been developed using omics and clinical data, but the most impactful approach is based on NanoString technology. However, tumor polyclonality-related time and space heterogeneity remain the biggest obstacles to proper molecular subtype verification. Thus, not only can different subtypes be detected in the primary tumor, but one subtype can be replaced with another subtype with time in a relapsed tumor or metastasis. Based on this data, clinical trials can be planned by recruiting precise molecular subtypes to achieve maximum efficacy. In addition, molecular subtyping appears to be an accurate tool for determining HGSC prognosis. In this chapter, the authors cover each HGSC molecular subtype in detail, with the most accurate verification approach and potential clinical applications.