To determine the relative accuracy of CT or MR imaging in the detection of inoperable tumor sites prior to cytoreductive surgery in a large series of patients with newly diagnosed primary epithelial ovarian cancer. One hundred thirty- seven women with newly diagnosed primary epithelial ovarian cancer underwent CT (n = 91) or MR imaging (n = 46) prior to cytoreductive surgery. The following imaging criteria were used to identify inoperable tumor sites: (1) peritoneal implants greater than 2 cm in maximum diameter in the porta hepatis, intersegmental fissure, gall bladder fossa, subphrenic space, gastrohepatic ligament, gastrosplenic ligament, lesser sac, or root of the small bowel mesentery; (2) retroperitoneal adenopathy greater than 2 cm in maximum diameter above the renal hila; (3) hepatic metastases or abdominal wall invasion. Imaging results were compared with operability at surgery. Cytoreductive surgery was suboptimal in 21 of the 137 (15% ) patients. Sixteen of these patients had inoperable tumor on preoperative imaging, while one additional patient had apparently inoperable tumor on imaging but was optimally debulked at surgery. The sensitivity, specificity, positive predictive value, and negative predictive value of preoperative imaging for the prediction of suboptimal debulking were 76% (16/21), 99% (115/116), 94% (16/17), and 96% (115/120), respectively. CT and MR imaging were equally effective (P = 1.0) in the detection of inoperable tumor. Preoperative CT and MR imaging are equally accurate in the detection of inoperable tumor and the prediction of suboptimal debulking in newly diagnosed epithelial ovarian cancer. This suggests imaging may help select patients who might be more appropriately managed by neoadjuvant chemotherapy. To determine the relative accuracy of CT or MR imaging in the detection of inoperable tumor sites prior to cytoreductive surgery in a large series of patients with newly diagnosed primary epithelial ovarian cancer. One hundred thirty- seven women with newly diagnosed primary epithelial ovarian cancer underwent CT (n = 91) or MR imaging (n = 46) prior to cytoreductive surgery. The following imaging criteria were used to identify inoperable tumor sites: (1) peritoneal implants greater than 2 cm in maximum diameter in the porta hepatis, intersegmental fissure, gall bladder fossa, subphrenic space, gastrohepatic ligament, gastrosplenic ligament, lesser sac, or root of the small bowel mesentery; (2) retroperitoneal adenopathy greater than 2 cm in maximum diameter above the renal hila; (3) hepatic metastases or abdominal wall invasion . Imaging results were compared with operability at surgery. Cytoreductive surgery was suboptimal in 21 of the 137 (15%) patients. Sixteen of these patients ha dnoperable tumor on preoperative imaging, while one additional patient had apparently inoperable tumor on imaging but was optimally debulked at surgery. The sensitivity, specificity, positive predictive value, and negative predictive value of preoperative imaging for the prediction of suboptimal debulking were 76% ( (P = 1.0) in the detection of inoperable tumor (16/21), 99% (115/116), 94% (16/17), and 96% (115/120) Preoperative CT and MR imaging are equal accurate in the detection of inoperable tumor and the prediction of suboptimal debulking in newly diagnosed epithelial ovarian cancer. This suggests imaging may help select patients who might be more likely managed more by neoadjuvant chemotherapy.