Objective: To explore the origin of ovarian high grade serous carcinoma (HGSC) through analysing the expression and significance of PAX8, PAX2, p53 and RAS in the ovary and fallopian tube of different types and grades of serous carcinoma. Methods: A total of 44 cases tissue samples of ovarian tumor including 34 malignant ovarian tumor and 10 normal normal tissue (as control group) were collected from the admitted patients in Affiliated Tumor Hospital of Guangxi Medical University from January 2015 to January 2016. Fallopian tube tissues were segmented in accordance with the fimbria, ampulla, isthmus and the corresponding ovarian tissues were by the side. There were 34 cases of patients with ovarian cancer including 29 cases of epithelial ovarian cancer (27 serous carcinoma, 1 mucinous carcinoma,1 endometrioid adenocarcinoma) and 5 non-epithelial ovarian cancer (sex cord-interstitial tumor). Among 27 cases of patients with ovarian serous cancer, there were 23 HGSC and 4 low-grade ovarian serous cancer (LGSC). One hundred fifty-three cases of samples were diagnosed as ovarian serous cancer by Shandong University Affiliated Qilu Hospital from 2005 to 2013 and these samples were made tissue microarray. (1) To analyze the expression and differences of PAX8, PAX2, p53 and RAS in the above tissues and tissue microarray from ovarian and tubal of HGSC and control women by immunohistochemistry methods. (2) To compare the expression levels of PAX8, PAX2, p53 and RAS in ovarian and fallopian tubes of ovarian cancer patients with different pathological types. (3) To analyze the correlations of tubal and ovarian tissue in PAX8, PAX2, p53 and RAS expression of HGSC. (4) To analyze the factors of the prognosis of ovarian serous cancer in tissue microarray by single factor analysis method. Results: (1) PAX8, PAX2, p53 and RAS expression was negative in normal ovarian epithelium of control group, but the expression of PAX8, PAX2, p53 and RAS were strongly positive brown in secrete cells of normal fallopian tube epithelium. (2) p53 and RAS expression of fallopian tube epithelium in the epithelial ovarian cancer group were significantly higher than those in the non-epithelial ovarian cancer groups (P<0.05), but the expression of PAX8 and PAX2 in fallopian tube and the expression of PAX8, PAX2, p53 and RAS in ovarian tissue was not statistically significant in the groups (P>0.05). PAX8, PAX2 and p53 expression of the ovarian in HGSC group were significantly higher than those in LGSC group (P<0.05), while the expression of RAS was lower in the ovarian of the high-grade group (P<0.05), while the expression of PAX8, PAX2, p53 and RAS in fallopian tube was not statistically significant in the groups (P>0.05). (3) There was a significantly positive correlation between fallopian tube and the corresponding ovary of HGSC in PAX8 and PAX2 expression (r=0.422, P=0.045; r=0.693, P=0.000), but not correlation in p53 and RAS expression (r=0.058, P=0.793; r= -0.190, P=0.384). (4) Univariate survival analysis showed that the progression free survival time in patients with ovarian serous cancer group was significantly correlated with the protein expression of PAX8, PAX2 and RAS (P<0.05), but there were not correlated with age, surgical staging, cell differentiation, lymph node metastasis and preoperative chemotherapy and p53 protein expression (P>0.05). The total survival time in patients with ovarian serous cancer group was significantly correlated with the protein expression of PAX8 (P<0.05), but there were not correlated with age,surgical staging, cell differentiation, lymph node metastasis and preoperative chemotherapy and the protein expression of PAX2, RAS and p53 (P>0.05). Conclusions: PAX8, PAX2, p53, RAS are of great significance for the study of origin of HGSC. HGSC may be derived from fallopian tube, but further investigation would be necessary to confirm this. PAX8, PAX2, p53, RAS could be expected to be used as predictors of survival prognosis in patients with ovarian serous cancer.