We report the clinicopathologic and immunohistochemical features in 8 patients with tubal or ovarian high-grade serous carcinoma that was present in uterine samples, in which there was the potential for clinical and morphologic misinterpretation as a primary uterine lesion before hysterectomy/bilateral salpingo-oophorectomy. Patients ranged in age from 45 to 70 yr (mean, 57 yr). The initial presentation was variable, ranging from incidental findings on routine Pap smears to pleural effusion. During the preoperative clinical investigation, 7 of 8 patients did not have evidence of an adnexal tumor based on physical examination and radiologic imaging, and serum CA-125 levels were normal to low in 4 of 5 patients. Six patients required multiple rounds of uterine samples, and the preoperative uterine specimens that contained lesional tissue and were available for rereview in all 8 patients included endometrial biopsies/curettages (n=6), endocervical curettages (n=3), Pap smears (n=2), and a hysteroscopic myomectomy specimen (n=1). The amount of carcinoma in these specimens was typically scanty. The lesions in most cases were characterized by detached and minute epithelial clusters, small papillae, and/or individual cells. The constituent glandular cells exhibited notable atypia. Psammoma bodies were identified in only 2 cases. Immunostains for WT-1 were positive in 3 of 4 preoperative specimens. All patients ultimately underwent a hysterectomy/bilateral salpingo-oophorectomy, which revealed an invasive high-grade serous carcinoma of tubal (n=6) or ovarian (n=2) origin. The mean/median tumor size was 3.2/1.7 cm. Transtubal spread was considered the most likely mechanism resulting in tubal/ovarian carcinoma being found in the preoperative uterine samples. These findings highlight the deceptive clinical features of some tubal/ovarian high-grade serous carcinomas, and demonstrate that small and clinically undetectable adnexal high-grade serous carcinomas can initially present in uterine biopsies/curettages. To guide clinical evaluation more accurately and prevent histologic misdiagnosis/misclassification, a possible adnexal origin should be considered in the differential diagnosis of small, detached, and markedly atypical glandular fragments in endometrial or cervical specimens, and immunohistochemical staining for WT-1 is recommended in this setting.