Although grading has been demonstrated to be an important prognostic factor in ovarian serous carcinoma, there is no system universally used to perform this task. A few years ago, we proposed a two-tier system for grading ovarian serous carcinoma that is based primarily on the assessment of nuclear atypia (uniformity vs. pleomorphism) in the worst area of the tumor. Tumor grade in this two-tier system is correlated with survival. After being used by numerous pathologists and trainees at The University of Texas M.D. Anderson Cancer Center (MDACC) for 15 years, we have observed that this system is user-friendly and reproducible. We undertook this study to evaluate the interobserver and intraobserver variability among a group of 7 gynecologic pathologists and 2 general surgical pathologists using this grading system. A total of 80 cases of ovarian serous carcinoma, 40 low-grade and 40 high-grade, were circulated twice among these pathologists. Slides with examples of low-grade and high-grade serous carcinoma were sent with the unknowns. A website was used to provide diagnostic criteria, images of examples of ovarian low-grade and high-grade carcinoma, and a log form to facilitate data entry. Statistical analysis demonstrated an overall kappa statistic among the different observers of 0.909. The intergrader kappa's ranged from 0.717 to 1.000 in the first round of the review and from 0.701 to 1.000 in the second round. Eight of the participants had an intragrader kappa ranging from 0.775 to 1.000 (excellent agreement), whereas a single participant had an intragrader kappa of 0.725 (good agreement). This study demonstrates that the two-tier grading system (the MDACC grading system) for ovarian serous carcinoma on the basis of the assessment of nuclear atypia is easy to learn and is highly reproducible. These findings would support its universal use, which would be beneficial for the standardization of clinical trials and protocols, thus facilitating the understanding of this disease and investigation into the treatment of patients affected by these tumors.