PURPOSE: The purpose of this study is to evaluate the accuracy of differentiation between benign and malignant clustered microcalcifications without mass on mammogram. MATERIAL AND METHODS: Fourty six mammogramsof 44 patients showing clustered microcalcifications without mass were interpreted blindly by five independent observers majoring in breast imaging from different institutions. Twenty two were malignant (10 infiltratingductal carcinomas, 12 intraductal carcinomas) and 24 were benign (all fibrocystic disease). The observers judgebenignancy or malignancy of microcalcifications. The authors assess the accuracy of differential diagnosis of clustered microcalcifications. RESULT: Of 24 cases proved benign microcalcifications, five radiologists correctly interpreted 20 on average as benign and of malignant 22 cases, 16 on average were correctly interpreted asmalignant. The diagnostic accuracy of malignant microcalcifications was 71.8% on average(63.6%-81.8%) and the diagnostic accuracy for benign microcalcifications was 83% on average(71% - 92%). It was 9 among total 46 cases that were misinterpreted by more than three radiologists. Among these 9 cases, malignant microcalcifications thathad been misinterpreted as benign were seven, benign microcalcifications misinterpreted as malignant were two. CONCLUSION: The diagnostic accuracy of clustered malignant microcalcifications(71.8%) without mass on mammogramwas lower than that of benign microcalcifications(83.3%). So, in case of suspected malignant microcalcification onmammogram, it is preferable that along with magnification view, histopathologic confirmation by core biopsy mustbe obtained.