Review of a 20-month experience with 241 peritoneal washes performed on 191 patients showed that the use of these specimens has expanded greatly. Of the 19 patients with neoplastic cells in their peritoneal washing cytology specimens, 12 had primary ovarian neoplasms, 4 had primary uterine cervical neoplasms, 2 had primary endometrial neoplasms, and 1 had mammary carcinoma metastatic to the ovary. Gynecologic oncologists at this institution are now routinely obtaining peritoneal washing cytology specimens whenever there is intraabdominal surgery on patients known to have or suspected of having a pelvic neoplasm. The following criteria were found to be essential to the accurate evaluation of these specimens: (1) cells considered to be malignant should be present both singly and in groups and should be malignant by the usual cytologic criteria, (2) the patients must have or be known to have had a neoplasm whose cells are similar to those in the washing specimen, and (3) the cells considered to be neoplastic must be different from and not confused with reactive mesothelial cells. The last criterion is important because the peritoneal lavage traumatically removes mesothelium, which can appear atypical. These criteria make the cytologic interpretation of most peritoneal washing specimens straightforward; interesting diagnostic problems occur, however, including the evaluation of neoplasms of borderline malignancy, those "spilled" during surgery and second neoplasms found by peritoneal washing cytology.