Sentinel lymph node biopsy (SLNB) has become the standard of care in most centers for axillary staging in patients with early breast cancer. Multiple radioactive nodes are often identified at surgery. The finding of multiple sentinel lymph nodes (SLNs) has been shown to be associated with lower rates of false-negative results in the SLNB procedure, hence the importance of removing and examining all SLNs. Often preoperative lymphatic mapping (PLM) is performed prior to surgery. In this study we examined whether the exact number of SLNs identified during surgery can be accurately predicted by PLM. During the years 2001-2004, 155 patients underwent both PLM and a SLNB in our breast unit. During surgery, an attempt was made to remove all radioactive nodes. The number of axillary radioactive foci found on PLM was compared with the number of radioactive nodes identified during surgery. The average number of sentinel nodes harvested was 2.3 (range 1-9). The average number of radioactive foci identified on PLM was 1.8 (range 0-5). Of the 155 patients, the number of sentinel nodes retrieved in surgery was greater than that found in preoperative mapping in 65 patients (41.9%), equal to that found in preoperative mapping in 60 patients (38.7%), and less than that found in preoperative mapping in 30 patients (19.4%). Thus in most patients, the number of SLNs found on PLM did not reflect the number of SLNs found intraoperatively. Therefore, even when the number of nodes identified on PLM has been reached in surgery, a meticulous search for additional nodes should still be carried out. The number of hot spots in preoperative mapping should serve as a rough indicator of the smallest number of nodes the surgeon should attempt to resect, but not the exact number of nodes expected to be found.