One hundred and fifty two patients presenting with palpable primary breast cancer were studied to evaluate accuracy of clinical assessment, mammography, and ultrasonography in measuring tumor size. The clinical, mammographic, and ultrasonographic diameter of tumor size was compared to histological diameters. The histological diameter of tumors was 27.0±12.9 mm. The average diameter of tumor on clinical assessment was 33.8±13.1 mm and its correlation coefficient to the histological size was 0.73. The average size on the mammography was 21.4±9.0 mm and its correlation coefficient to the histological size is 0.71. The tumor size on the ultrasonography in 22 patients was is 22.3±10.4 mm. The sonographic measurement demonstrated the highest correlation coefficient (0.83). The clinical assessment overestimated the tumor size, but the mammographic and sonographic measurement underestimated it compared with histological size. The combined measurement with clinical and ultrasonographic methods slightly improved the con-elation with histological size (0.86). In conclusion, ultrasonography could be useful method to estimate tumor size preoperatively. And when it is combined with clinical assessment, it is the most reliable and accurate technique.