As compared to conventional axillary dissection, the sentinel node technique is accompanied by reduced morbidity and shorter hospital stay. Based on available data, the use of this technique does not seem to yield higher rates of axillary recurrence. A combination of both radioisotope detection and blue dye increases the identification rate, while also reducing false-negative rate. Surgical results are optimized when preoperative lymphoscintigraphy mapping is obtained in addition to peroperative probe detection. Considering the site of injection, the subareolar injection can be easy to apply even in case of non-palpable tumours, and gives higher count rates. However, the intraparenchymal, peritumoral, injection is necessary to evidence cases of extra-axillary drainage (internal mammary, infra- or supraclavicular) that is present in about 20% of patients. With the advent of hybrid cameras (SPECT-CT), the topography of these extra-axillary nodes can be given with high precision. Use of the sentinel node technique has been accompanied by an increase in the percent of patients with node involvement, due to an increased detection of micrometastases inferior or equal to 2 mm. Following an overview of basic principles, and of the main results with the sentinel node technique we focus the discussion on several points that are still open to debate, such as: 1) which group of patients can benefit from the sentinel node technique? 2) What is the optimal methodology? 3) What is the prognostic significance of micrometastases and of isolated tumour cells? 4) What attention should be given to extra-axillary drainage?