Abstract: Objective To analyze the pathological underestimation of breast lesions by ultrasound-guided 14G type core needle biopsy(CNB), and discuss the causes of different pathological underestimation. Methods A total of 758 breast lesions［radiographic examination showed lesions were the fourth classification or above by American College of Radiology Breast Imaging Report and Data System(BI-RADS)］ were conducted by ultrasound-guided CNB, the results of CNB were compared with subsequent surgical pathology results, and the pathological underestimation conditions of breast lesions［including underestimation of papillary lesion, escalation of benign to malignant lesions(except for papillary lesion and phylloides sarcoma), benign-lobular carcinoma in situ(LCIS, special type of escalation of benign to malignant grading), intraductal carcinoma(DCIS)-microinvasive carcinoma, and DCIS-invasive carcinoma］ were summarized, and the reasons were analyzed. Results Overall underestimation rate of CNB was 8.3%(63/758). Underestimation rates was 17.4%(4/23)in papillary lesion, 65.6%(21/32)in escalation of benign to malignant grading lesions(including 1 lesion of escalation of benign to LCIS), 26.3%(15/57)in DCIS-microinvasive carcinoma, and 40.4%(23/57)in DCIS-invasive carcinoma. Conclusion The results of CNB are mostly consistent with final pathological results, but underestimation conditions stills occur(papillocarcinoma can be underestimated, thus tumor excision for biopsy in women aged above 50 years who have suspected papillary lesion is recommended. As for patients with escalation of benign to malignant lesions, intraductal papilloma, lobular tumor and lobular carcinoma in situ, immunohistological examination and open surgical biopsy - are needed for further diagnosis).