Study Objective: To estimate the prevalence of nodal involvement according to various disease characteristics in order to assess the prognostic advantages to have nodal dissection in apparent early-stage epithelial ovarian cancer (eEOC). Design: Retrospective study. Setting: Gynecologic oncology referral center. Patients: Data of consecutive patients undergoing comprehensive staging for eEOC were retrospectively evaluated. Logistic regression and a nomogram-based analysis were used to assess the risk of nodal involvement. Intervention: Retroperitoneal staging including pelvic and para-aortic node dissection. Measurements and Main Results: Overall, 377 patients were included. All patients had nodal dissection including pelvic and para-aortic lymphadenectomy in 366 and 370 cases, respectively. Forty-four (11.7%) patients were upstaged due to nodal involvement. Pelvic and para-aortic nodal metastases were observed in 32/366 (8.7 %) and 42/370 (11.3%) patients, respectively. Nodal involvement was observed in 46/136 (33.8%), 8/24 (33.3%), 15/94 (15.9%), 4/42 (9.5%) and 1/81 (1.2%) patients with serous, undiffentiated, endometrioid, clear cell, and mucinous histology (p < .001). Via multivariate analysis, we observed that poor differentiated tumor, (FIGO grade 3), serous histology and bilateral tumors wereindependently associated with both pelvic and para-aortic nodal involvement (p < .01). The risk of presence of nodal involvement was calcutaed using a nomogram. Conclusion: Our data suggested that FIGO grade 3, serous and bilateral eEOC are at high risk of having disease harboring in the lymph nodes. After receiving external validation, our data will help to identify patients deserving comprehensive retroperitoneal staging.