Several articles have been published recently that discuss the role of electron microscopy in the diagnosis and study of gynecologic neoplasms. It becomes apparent from those works and the review just presented that, although an ultrastructural study is not necessary for reaching a diagnosis of many of these tumors, it may be necessary or supportive in identifying the more poorly differentiated ones. Furthermore, electron microscopy is valuable in providing evidence for the histogenesis of some of these neoplasms. Unfortunately for the pathologist, a certain level of morphologic differentiation (and an absence of metaplasia) in a cell is usually necessary for these goals to be achieved. For example, an adenomatoid tumor (see the article by Dr. Srigley, Mr. Toth, and Mr. Edwards in this issue) of the fallopian tube can readily be accepted as being composed of mesothelial cells, because both the neoplastic cells and normal mesothelial cells have the same highly differentiated features of long, slender microvilli, prominent intercellular junctions, and many microfilaments. On the other hand, there is very little resemblance between the granulosa cells of a granulosa-cell tumor and mature mesothelial cells. Thus, if one of the theories of histogenesis of granulosa cells were correct--namely, that they are derived ultimately from mesothelial lining--the ultrastructural evidence would rest on recognizing a similarity between the two types of cells at an earlier stage of differentiation. The neoplastic granulosa cell has differentiated along a separate, specialized line in which the ultrastructural resemblance to the parent cell is partly, if not completely, lost. Another example of the type of information that electron microscopy can provide is in relation to the common epithelial tumors. There is good evidence that the serous tumors in this group arise from mesothelium, although ultrastructurally their differentiation has veered from a mesothelial direction to one in which the cells have a complement of organelles related to secretory activity. Paradoxically, the mucinous cystic tumors, which have been classified traditionally as tumors of surface epithelial origin, are now thought to be of germ-cell origin in some cases, as examples of monophyletic teratomas. The ultrastructural evidence for this conclusion rests on the presence of anchoring filaments in the microvilli of the neoplastic cells, similar to those of normal intestinal epithelium, and on an admixture of various types of gastrointestinal cells, including those that contain dense-core granules (argentaffin cells).(ABSTRACT TRUNCATED AT 400 WORDS)