Although squamous differentiation within endometrial carcinomas has long been recognized by pathologists, its biologic significance has been the subject of continued debate. While some authors have found a worsened prognosis for women who have tumors with squamous elements, others have reported the prognosis to be better than conventional endometrial adenocarcinomas. Persisting confusion and disagreement about the use of the terms adenoacanthoma and adenosquamous carcinoma have complicated the issue. In this article we review the literature on the pathogenesis of squamous differentiation in the endometrium and discuss the histologic features, prevalence, and biologic behavior of adenocarcinoma with squamous differentiation. We conclude that keratin is a constituent of normal and neoplastic endometrial epithelial cells, and that overt squamous differentiation occurs by a mechanism that is currently unknown. Squamous differentiation is present in about 25% of endometrial adenocarcinomas, a frequency that appears to have been constant for the past 50 years. The squamous component of endometrial carcinomas may histologically appear benign, malignant, or indeterminant, and in the majority of instances closely parallels the differentiation of the glandular component. Endometrial adenocarcinomas with malignant-appearing squamous elements usually have poorly differentiated glandular components and have a prognosis identical to that of poorly differentiated adenocarcinoma without squamous differentiation. Endometrial adenocarcinomas with benign-appearing squamous elements are usually associated with well-differentiated glandular components and have a prognosis identical to that of typical well-differentiated adenocarcinoma. Some endometrial adenocarcinomas contain foci of squamous differentiation that appear neither clearly benign nor malignant. These often are associated with moderately differentiated glandular components; the prognosis for these women is not yet clearly defined. At the present time we are unable to attribute any prognostic significance to the presence of squamous differentiation in endometrial carcinomas. Because of the confusion and semantic arguments that revolve about the use of the terms adenoacanthoma and adenosquamous carcinomas, we recommend that those terms be abandoned and be replaced by the single term adenocarcinoma with squamous differentiation. As for any other endometrial tumor, the pathologist should provide information on histological grade, depth of myometrial invasion, and presence of vascular involvement or spread to the cervix in order to guide the gynecologist in determining appropriate therapy.