Sentinel lymph node (SLN) mapping has revolutionized the way we stage breast cancer. A blue dye technique (BD) and the use of a radiotracer with the assistance of a gamma-detecting probe (GDP) have been used for the identification of the sentinel nodes. Some groups have suggested that only one technique is necessary. The reported false negative rates have been 0 to 12 per cent and success rates as low as 65 per cent. We have prospectively evaluated these techniques and have used both for the identification of the SLN. Ten surgeons participated in this study. From April 1998 through May 1999, 58 patients underwent SLN mapping followed by an axillary lymph node dissection. After the injection of 0.3 to 1.96 mCi of filtered sulfur colloid diluted to 4 mL all patients had preoperative lymphoscintigraphy. Five minutes before surgery 3 to 5 mL of isosulfan blue was injected around the tumor or tumor bed. Even though preoperative lymphoscintigraphy identified an SLN in 35 patients (63%) successful intraoperative detection of an SLN was possible using both techniques in 53 patients (91%). The SLN was detected by the BD and the GDP in 37 (65%) and 45 (80%) respectively. Nineteen patients (33%) were positive for metastatic disease in the axilla. Twenty-two (19%) of 113 SLNs removed were positive for disease. All cases of metastatic disease in the axilla were detected by the mapping technique. False negative rate was 0 per cent. In 11 patients the only positive node was the sentinel node (58%). Furthermore six (32%) patients were upstaged by the use of immunostains for cytokeratin. Twenty-two positive SLNs were detected in the 19 patients. The positive lymph node was identified only by BD in four patients (21%), only by GDP in six patients (31%), and by both techniques in nine patients (47%). We conclude that if only one technique had been used the false negative rate could have been as high as 32 per cent. Both techniques must be used to obtain a low false negative rate and high yield in the identification of the SLN.