PURPOSE: This study was performed to find the adequate number of removed lymph nodes to achieve an acceptable false-negative rate when performing sentinel lymph node biopsy for breast cancer. METHODS: A total of 179 sentinel node biopsies combined with conventional axillary lymph node dissection for breast cancer were performed between November 2003 and June 2007. RESULTS: The overall identification rate of sentinel lymph node and the false negative rate of the biopsy were 95.0% and 8.1%, respectively. Yet the false negative rate of the biopsy was lowered as the number of the removed nodes was increased. Especially, the false negative rate was 0% when more than 4 lymph nodes were removed. CONCLUSION: We recommend that four lymph nodes should be removed to obtain accurate results in sentinel node biopsy for breast cancer.