In this study, we aimed to determine whether the presence of poorly differentiated histologic components in ovarian clear cell adenocarcinoma (CCA) affects patient prognosis. Pathologic slides from 159 patients with CCA were studied, and the tumors were classified as Por(+) in the event of poorly differentiated histology; that is, if solid masses or cords, or individual infiltrating tumor cells with little or no glandular/papillary differentiation were present in >5% of the tumor area examined. All other tumors were classified as Por(-). The prognostic value and interobserver reproducibility of this assignment were analyzed. The agreement level in the assignment between 2 pathologists was 93.7% (κ=0.86). After a consensus was reached, 53 (33%) and 106 (67%) tumors were classified as Por(+) and Por(-), respectively. Patients with Por(+) tumors showed a significantly worse outcome than those with Por(-) tumors, both in the early stages (stages I/II, 5-year survival rate 53.9% vs. 96.3%, P<0.0001 by log-rank test) and advanced stages (stages III/IV, 5-year survival rate 26.5% vs. 49.2%, P<0.001 by generalized Wilcoxon test). Por(-) tumors showed an effective response to postoperative platinum-based first-line chemotherapy more frequently compared with Por(+) tumors (48% vs. 14%, P=0.042). The Por(+) tumor was found to be an independent prognostic factor for survival irrespective of the clinical stage or presence of residual tumor after the initial surgery. These results suggest that the tumor with a poorly differentiated histology is an adverse prognostic subgroup in ovarian CCA. On the basis of the prognostic impact and interobserver reproducibility, the present binary classification system for CCAs was deemed to be highly superior to the compared conventional histologic grading system.