Ovarian carcinoma is the most lethal cancer among all gynecological malignancies due to recurrence through chemoresistance. The aim of the present study was to identify new biomarkers to predict chemoresistance and prognosis in ovarian carcinomas. The mRNA expression by qRT-PCR was examined in 60 ovarian serous carcinomas for selected genes from the screening by PCR array focusing on apoptosis, epithelial-to-mesenchymal transition and cancer pathways. The clinical impact was assessed by analyzing the correlation between gene expression and clinicopathological variables. Further validation with immunohistochemistry was performed with 75 cases of serous carcinomas. The chemoresistance was significantly associated with high expression of FOS, GSC, SNAI1, TERT and TNFRSF10D, and low expression of CDKN1A, TNFRSF10A, TNFRSF10C and TRAF1 (p<0.05, t-test). Low expression of TRAF1 and high expression of E2F1, FOS, TERT and GSC were significantly associated with advanced clinical stage (p<0.05, χ2-test). Lymph node metastasis was significantly associated with high expression of GSC. The upregulation group of TERT, GSC, NOTCH1 and SNAI1, and downregulation group of TRAF1 were significantly associated with poor overall survival (p<0.05, log-rank test). On further validation with immunohistochemistry, overexpression of goosecoid homeobox (GSC) was associated with poor overall survival. The results suggest that GSC is the most potential biomarker of drug response and poor prognosis in ovarian serous carcinomas.