A retrospective study of 567 patients treated for surgical stage I endometrial cancer was undertaken to resolve the controversy over the significance of malignant peritoneal cytology findings in early-stage disease. Twenty-eight women (4.9%) had peritoneal cytology positive for malignant cells. Comparisons were made between the groups with positive and negative cytology. Subgroups used in analysis included stage according to the International Federation of Gynecology and Obstetrics, treatment regimen, histology, grade, depth of myometrial invasion, and cervical Papanicolaou smear results. Cervical smear status was the only subgroup in which a statistically significant difference was found, with the positive peritoneal cytology patients having a higher incidence of positive Papanicolaou smears (P = .01). Forty-nine women (8.6%) developed recurrent tumor, 7% of the negative-cytology group and 32% of the positive-cytology group (P = .0002). The progression-free survival rate was lowered significantly by positive peritoneal cytology (P less than .0001); patients with negative peritoneal cytology had a significantly better 5-year survival rate, 96 versus 84% (P = .0001). When multivariate analysis was performed on the 477 cases that had no missing values, peritoneal cytology remained significant for both survival rate (P = .01) and progression-free interval (P = .002). Positive peritoneal cytology is a poor prognostic factor for patients with surgical stage I endometrial cancer.