Management of the axilla in patients with operable breast cancer is still one of the most controversial areas in clinical oncology. The best procedure to examine the lymph nodes is still standard axillary lymph node dissection; nevertheless, the morbidity associated with this procedure is well known. Based on these considerations, it is important for progress in the treatment of operable breast cancer that strategies are found that permit a less invasive method of axillary sampling which does not impair the patient's quality of life. The technique of sentinel lymph node (SLN) biopsy has recently been proposed for this purpose, with very important results. SLN has now become routine practice in the surgical management of breast cancer, and in many institutions patients with a negative SLN biopsy are spared axillary dissection, while those with a positive SLN biopsy are submitted to axillary node dissection. The good accuracy of SLN biopsy represents a significant advance in the management of primary breast cancer; however, false negative axillary results can occur in a variable percentage of patients, and the contribution of the SLN procedure to the detection of metastases in the internal mammary and supraclavicular lymph nodes is not clear. Among the recently developed imaging modalities, positron emission tomography (PET) with (18)F-fluorodeoxyglucose (FDG) has in particular been applied to the study of lymph node metastases in cancer patients. Several clinical studies have been carried out to evaluate the accuracy of PET in the axillary staging of operable primary breast cancer. These studies have sometimes provided conflicting results, either supporting the possibility of using FDG-PET to select patients who need axillary dissection or questioning whether FDG-PET can accurately assess the axillary status in primary breast cancer. All the limitations and the advantages of FDG-PET are discussed in this paper, by examining the performance of scanner technology and the possible causes of the false negative results. In the experience of the authors, comparing FDG-PET with SLN biopsy in the same series of patients, the results seem to indicate that the lower sensitivity of PET is restricted to micrometastases. Of course, this limitation of PET has to be analysed in relation to the importance of such small axillary metastases for the outcome of patients with breast cancer. The added value offered by PET in breast cancer staging in comparison with intraoperative detection of the sentinel node lies in the fact that FDG-PET is a non-invasive procedure that allows, within a single examination, the biological characterisation of breast cancer and viewing of the entire body.