Although infrequently encountered, the diagnosis of ovarian high-grade endometrioid carcinoma remains a diagnostic challenge with potential consequences for targeted therapies and genetic counselling. We studied the clinical, morphologic, and immunohistochemical features of ovarian high-grade endometrioid carcinomas and their diagnostic reproducibility compared with tuboovarian high-grade serous carcinomas. Thirty cases confirmed as International Federation of Gynecology and Obstetrics grade 3 endometrioid carcinomas were identified from 182 ovarian endometrioid carcinomas diagnosed in Alberta, Canada, between 1978 and 2010, from the population-based Alberta Ovarian Tumor Types cohort. Cases of lower grade endometrioid and high-grade serous carcinoma served for comparison. Ten immunohistochemical markers were assessed on tissue microarrays. Clinical data were abstracted and survival analyses performed using Cox regression. Interobserver reproducibility for histologic type was assessed using 1 representative hematoxylin and eosin–stained slide from 25 randomly selected grade 3 endometrioid carcinomas and 25 high-grade serous carcinomas. Histotype was independently assigned by 5 pathologists initially blinded to immunohistochemical WT1/p53 status, with subsequent reassessment unblinded to WT1/p53 status. Patients diagnosed with grade 3 endometrioid carcinoma had a significantly longer survival compared with high-grade serous carcinoma in univariate analysis (hazard ratio [HR]=0.34, 95% confidence interval [CI]=0.16-0.67, P=0.0012) but not after adjusting for age, stage, treatment center, and residual tumor (HR=1.01, 95% CI=0.43-2.16, P=0.98). Grade 3 endometrioid carcinoma cases (N=30) were identical to grade 2 endometrioid carcinoma cases (N=23) with respect to survival in univariate analysis (HR=1.07, 95% CI=0.39-3.21, P=0.89) and immunohistochemical profile. Using histomorphology alone, interobserver agreement for the diagnosis of grade 3 endometrioid or high-grade serous carcinoma was 69%, which significantly increased (P<0.0001) to 96% agreement with the knowledge of WT1/p53 status. Our data support the diagnostic value of WT1/p53 status in differentiating between grade 3 endometrioid carcinoma and high-grade serous carcinoma. However, grade 3 and grade 2 endometrioid carcinomas showed no differences in immunophenotype or clinical parameters, suggesting that they could be combined into a single group.