From July 1975 to April 1983, 237 patients had primary treatment for endometrial cancer at the Long Island Jewish-Hillside Medical Center. Included in this study were 74 of these patients with Stage I and 20 with Stage II endometrial carcinoma who underwent laparotomy without preoperative radiation. The purpose of the study was to determine the prevalence of extrauterine spread in endometrial carcinoma clinically confined to the uterus and to correlate risk variables with this spread. The parameters assessed were retroperitoneal nodal metastases, adnexal involvement, peritoneal implants and peritoneal cytology. The overall prevalence of extrauterine spread was 23.4% (Stage I, 18.9%; Stage II, 40.0%). The rate of nodal metastasis, adnexal involvement, peritoneal implant, and positive peritoneal cytology were 18.7, 7.4, 4.3, and 8.5%, respectively. No positive relationship was demonstrated between surface spread and risk variables. There was positive correlation between surface spread and peritoneal cytology (87.5%). Direct correlations were found between positive nodes and tumor growth over more than one-third of the endometrial surface (P less than 0.001), gross cervical involvement (P less than 0.001), deep myometrial invasion (p less than 0.001), length of uterine cavity, grade 3 tumor, papillary adenocarcinoma (40%), and stage of disease. Five-year survival rate of Stage I and Stage II in this small series was 77.8 and 55.6%. Complications of 16 radical hysterectomies in Stage II were minimal and transient. Because of frequent extrauterine spread in endometrial carcinoma clinically confined to the uterus, and exploratory laparotomy and peritoneal cytology may be desirable in Stage I and II disease before definitive treatment.