The FIGO grading for endometrial endometrioid carcinomas is widely accepted. In 2000, a novel binary architectural grading system was suggested that divided endometrioid carcinomas into low- and high-grade tumors. We aimed to evaluate the interobserver reproducibility of the FIGO, the architectural binary, and nuclear grading systems and the correlation between these grading systems and pathologic prognostic factors for endometrial endometrioid carcinoma. Eighty-eight endometrial endometrioid carcinomas from hysterectomy specimens were reevaluated by two pathologists independently. Kappa values for the FIGO, the binary, and the nuclear grading systems were 0.65, 0.67, and 0.09, respectively. The reproducibility of the FIGO and the binary grading systems was similar and substantial. FIGO grade 1 (60.2%) patients were comparable to binary low-grade (63.6%) patients. Most of the FIGO grade 3 (83%) patients were binary high grade. FIGO grade 2 patients were distributed between binary low and high grades. The FIGO grade 1 and 2 cases judged to be of binary high grade had deep myometrial invasion, and more cases had vascular invasion in comparison with FIGO grade 1 and 2 cases judged to be of binary low grade. In uni- and multivariate analyses, both grading systems, depth of myometrial invasion, vascular invasion, cervical involvement, and stage had no effect on overall survival. But binary high grade and vascular involvement are adverse prognostic factors on recurrence-free survival. Binary high-grade patients can be assigned as high-risk patients.