PURPOSE: The purpose of this study was to compare the diagnostic performance of ultrasound (US) elastography and conventional B-mode US for discrimination between benign and malignant breast lesions. MATERIALS AND METHODS: During a 13-month period, 277 women with 335 sonographically visible breast lesions who were scheduled to undergo biopsy were examined with US elastography. Elastographic findings were classified as benign or malignant based on the area ratio, with 1.00 as the threshold. Findings on conventional B-mode US were classified according to the BI-RADS category, as follows: lesions of BI-RADS categories 2 and 3 were considered benign, while those in categories 4 and 5 were considered malignant. Statistical analysis included sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), accuracy, and ROC curve analysis for comparison of the diagnostic performance of US elastography and conventional B-mode US. RESULTS: Of the 335 breast lesions, 85 (25.4%) showed malignancy on pathology. Findings on B-mode US showed malignancy in 264 (78.8%) and elastographic findings showed malignancy in 102 (30.4%). The sensitivity, specificity, PPV, NPV, and accuracy of B-mode US and elastography were 98.8%, 28.0%, 31.8%, 98.6%, and 79.4% and 69.4%, 81.2%, 57.8%, 88.8%, and 79.4%, respectively. Elastography showed significantly higher specificity and PPV and lower sensitivity and NPV, compared with B-mode US (p < 0.001). The area under the ROC curve (AUC value) was 0.761 for elastography, and 0.634 for B-mode US (p < 0.001). CONCLUSIONS: US elastography can improve specificity and PPV of B-mode US, but with significant sacrifice of sensitivity and NPV. Therefore, US elastography may complement B-mode US for differentiation of breast masses.