The reported experience with early-stage (FIGO stage I/II) ovarian carcinoma (OC) is limited given that the majority of women with OC are diagnosed at an advanced stage. There has not been an extensive review of these tumors, and since the pathologic criteria differentiating invasive and borderline tumors have evolved over time, the issue of whether a proportion of these tumors should be reclassified has not been addressed. We identified patients with stage I/II invasive OC who underwent primary surgical management at Memorial Sloan-Kettering Cancer Center from 1980 to 2000. Patients known to have a BRCA mutation or a family history of breast/ovarian cancer were excluded. Hematoxylin and eosin slide review, blinded to clinical outcomes, using current diagnostic criteria for ovarian carcinomas and borderline ovarian tumors, was performed. Progression-free survival (PFS) and disease-specific survival (DSS) were estimated and compared. Hematoxylin and eosin slides were reviewed for 140 of the 145 patients identified. The diagnosis was changed to borderline (low malignant potential) in 41 cases (29.3%). Twenty-nine (70.7%) of 41 changes in diagnosis involved endometrioid and mucinous tumors. This was attributable to the application of recently revised criteria for distinguishing borderline tumors from carcinomas. None of the originally diagnosed clear cell carcinomas was reclassified as borderline. The distribution of histologic subtypes among the 94 carcinomas included 26 serous (27.7%), 25 clear cell (26.6%), 22 endometrioid (23.4%), 10 mixed (10.6%), 6 mucinous (6.4%), 2 malignant Brenner (2.1%), and 3 adenocarcinomas, not otherwise specified (3.2%). Adjuvant therapy was given to 84 (89.4%) of the 94 patients with carcinomas. The 5-year PFS and DSS were significantly greater for the group of cases that was reclassified as borderline (4.5% vs. 26.2% progressed [P = 0.006]; 4.5% vs. 25.6% died [P = 0.003]). The 5-year PFS and DSS were significantly worse for carcinomas with a TP53 mutation (22.6% vs. 41.2% progressed [P = 0.04]; 21.7% vs. 24.7% died [P = 0.04]). There were no statistically significant differences in outcome between stages I versus II, tumor grades, clear cell histology versus other, and stage IC preoperative versus intraoperative rupture. We concluded that a large number of cases originally diagnosed as early-stage sporadic OC were borderline tumors. Clear cell histology does not confer a worse prognosis compared with other histologies. The presence of a TP53 mutation was an adverse prognostic indicator.