Cytologic examination is not a widely accepted method for the detection of endometrial lesions. Cervical smears may provide useful information about the endometrium in some cases; however, it is not a sensitive test, and differentiating abnormal endometrial cells from normal cells in a cervical smear is not always possible. On the other hand, direct intrauterine sampling by using the Tao brush (Cook Ob/Gyn, Spencer, IN) was introduced as a reliable method for detecting endometrial abnormalities. The main purpose of endometrial brush cytology is to diagnose endometrial hyperplasias and carcinomas. It also provides information about the endometrial phase, and has the potential to be used as a screening device for patients at high risk. [1][2][3][4] We planned a study to determine the accuracy of endometrial brush cytology processed by the ThinPrep method. The preliminary results are presented here.Endometrial samples were collected from 35 consecutive hysterectomy specimens by using the Tao brush. The brush, covered with the outer sheet, was inserted to the level of the fundus, and the outer sheet was pulled back. The brush was rotated 360°, the outer sheet was pushed back to the tip, and the device was removed. After pulling back the outer sheet, the brush was immersed into 20 ml of PreservCyt solution. The sheet was moved up and down in order to expose the adherent cells and tissue fragments to the solution. Slides were prepared from vials using the ThinPrep 2000 (Cytyc Corp.) automated slide preparation system and stained by the Papanicolaou method. Cytologic features were evaluated according to the previously described criteria of Tao. 4,5 The slides were interpreted without knowing the final histologic diagnosis, and the results were compared.The cytologic diagnosis was proliferative endometrium in 11 cases; all were confirmed by histology except one, which was a secretory endometrium. Seven cases were classified as secretory endometrium by cytology; the histologic diagnosis was also secretory endometrium in all of them. Cytologic findings were interpreted as atrophic in 12 cases: 1 was misclassified (the histologic diagnosis was proliferative endometrium), 3 showed basal endometrium in hysterectomy specimens, and in the other 8 cases, the cytologic diagnosis was confirmed by histology. Three cases represented cycling endometrium; however, they could not be classified as either proliferative or secretory. The final histologic diagnosis was proliferative in one case, and secretory in the others. Cytologic material was inadequate for interpretation in 2 cases (Table I).Straight tubules with nuclear crowding and overlapping characterized the proliferative phase. Glandular cells showed scant cytoplasm and ovoid nuclei, with relatively dense chromatin (Fig. 1). In the samples of the secretory phase, flat sheets were more common than tubules. The nuclei of glandular cells were larger and rounded, with fine chromatin and small nucleoli. They had moderate amounts of cytoplasm, with prominent perinuclear clearing. No o...