Objective：To investigate the expression of HOXA9 protein in ovarian serous tumors and its potential as a specific marker for early diagnosis and drug resistance in ovarian cancer. Methods：Immunohistochemical staining for HOXA9 was performed in 178 patients with serous ovarian carcinoma，in 72 patients with ovarian borderline tumors and in 69 patients with benign ovarian tumors. The correlations between high HOXA9 expression and the clinicopathological features of the ovarian carcinomas were evaluated by the χ2 test and Fisher's exact test. The overall survival（OS）rates were calculated by the Kaplan-Meier method，and the association between prognostic factors and patient survival was analyzed by the Cox proportional hazard model. Results：The positive rate of HOXA9 protein in serous ovarian cancer（75.8%，135/178）was significantly higher than that in benign serous tumors（34.8%，24/69）（P < 0.05）. Similarly，the strong positive expression rate of HOXA9 protein in serous ovarian cancer（59.6%，106/178）was significantly higher than that in benign serous tumors（11.6%，8/69）（P < 0.05）. The positive expression rate of HOXA9 in high grade（65.1%，71/109）ovarian cancer was significantly higher than that in lower grade（47.8%，33/69）（P=0.014）. Similarly，the positive expression rate of HOXA9 in metastatic ovarian cancer（72.6%，69/95）was significantly higher than that in ovarian cancer without metastasis（42.2%，35/83）（P < 0.001）. Regarding the FIGO clinical stage，strong positive HOXA9 expression was 69.1%（65/94）in advanced（ⅡB-ⅢC）ovarian cancers and only 46.4%（39/84）in early（stageⅠ-ⅡA）ovarian cancers. The difference was statistically significant（P=0.003）. The OS rate of patients with strong expression of HOXA9 was lower than that of patients with weak expression of HOXA9（Log-rank=43.765，P < 0.001）. Multivariate analysis showed that HOXA9 was an independent prognostic factor in patients with ovarian cancer. Conclusion：HOXA9 expression is strongly associated with grade and outcome in ovarian carcinoma，and may serve as a useful molecular marker for clinical management.