OBJECTIVES. The goal of this research was to determine the value of the direct endometrial brush cytology in the detection of premalignant and malignant lesions of endometrium. STUDY DESIGN. A retrospective cross-sectional comparative study. PARTICIPANS AND METHODS. The study included subjects who were subjected to the direct endometrial brush cytology at the Department of Clinical Cytology, Clinical Hospital Center Osijek, from 2002 to the end of 2015, provided that they were also subjected to the histological verification, over a period of 6 months after the cytological diagnosis. The study also included subjects with normal cytological findings without histological verification, but with clinical and ultrasound follow-up for at least three years. Exclusion criteria were inadequate specimens. Cytology diagnosis were benign lesions, hyperplasia, hyperplasia with atypias and endometrial carcinoma. Cytology brush samples were fixed in 95% alcohol and stained with Papanicolaou method. Histologic analysis was performed at the Department of Pathology and Forensic Medicine, Clinical Hospital Center Osijek. Histopathological analysis was done on formalin fixed, paraffin embedded samples obtained from fractionated curettage or hysterectomy resection material, according to the current WHO classification. RESULTS. The sensitivity of the uterobrush endometrial cytological examination method for benign lesions and hyperplasia was 90%. The specificity was high in detection of all diagnoses and it ranged from 87.1% to 99.3%. Positive predictive value was high for detecting benign lesions (98.4%) and malignant lesions (89.5%) but low for hyperplasia with atypias (86%). The total diagnostic accuracy of the differential cytology was 91.6%. CONCLUSION. The direct endometrial brush cytology is a reliable diagnostic method for differentiation between benign and malignant lesions and its use can minimize the need for other invasive and more expensive procedures. KEYWORDS. OBJECTIVES. The goal of this research was to determine the value of the direct endometrial brush cytology in the detection of premalignant and malignant lesions of endometrium. STUDY DESIGN. A retrospective cross-sectional comparative study. PARTICIPANS AND METHODS. The study included subjects who were subjected to the direct endometrial brush cytology at the Department of Clinical Cytology, Clinical Hospital Center Osijek, from 2002 to the end of 2015, provided that they were also subjected to the histological verification, over a period of 6 months after the cytological diagnosis. The study also included subjects with normal cytological findings without histological verification, but with clinical and ultrasound follow-up for at least three years. Exclusion criteria were inadequate specimens. Cytology diagnosis were benign lesions, hyperplasia, hyperplasia with atypias and endometrial carcinoma. Cytology brush samples were fixed in 95% alcohol and stained with Papanicolaou method. Histologic analysis was performed at the Department of Pathology and Forensic Medicine, Clinical Hospital Center Osijek. Histopathological analysis was done on formalin fixed, paraffin embedded samples obtained from fractionated curettage or hysterectomy resection material, according to the current WHO classification. RESULTS. The sensitivity of the uterobrush endometrial cytological examination method for benign lesions and hyperplasia was 90%. The specificity was high in detection of all diagnoses and it ranged from 87.1% to 99.3%. Positive predictive value was high for detecting benign lesions (98.4%) and malignant lesions (89.5%) but low for hyperplasia with atypias (86%). The total diagnostic accuracy of the differential cytology was 91.6%. CONCLUSION. The direct endometrial brush cytology is a reliable diagnostic method for differentiation between benign and malignant lesions and its use can minimize the need for other invasive and more expensive procedures.