Abstract Background: Large multi institutional studies have pointed that previous surgical resection of breast tumours before axillary sentinel node detection (ASLND) was the main criteria of failure of this technique. Screening campaigns provide small tumours and despite efforts to obtain a diagnosis of early breast cancer, this is not always obtained, due to small tumours or false negative results of micro biopsies. The aim of our series was to assess identification rates and false negative rates of ASLND after previous surgical resection of breast tumours. Material and Methods: In a prospective multi institutional setting (14 multidisciplinary teams), we have included patients with a previous breast tumour surgical resection for the diagnosis of infiltrative breast adenocarcinoma. Patients with only a core biopsy and no surgical removal of the tumor before axillary surgery were not included. Each patient underwent a secondary surgical procedure for ASLND and axillary lymphadenectomy, and sometimes a breast secondary surgical procedure for margins. ASLND was performed with the combined method, with blue dye and technetium. Pathology was performed with serial sectioning, eosin safron and immune histo chemistry (IHC). Results: From July 2006 to November 2011, 138 patients where included. The median tumor size was 9mm. Identification rate was 86% (118/138). A macrometastasis was found in 11 cases, in a sentinel node (9), or in a non sentinel node(2). False negative rate was 9% (1 false negative sentinel node with macrometastasis in non sentinel node from lymphadenectomy/11 cases with a macrometastasis in either a sentinel node or a non sentinel node). In 1 case a micrometastasis was found in a sentinel node through IHC, with a macrometastasis in a non sentinel node from lymphadenectomy. Without IHC or without the decision of performing a complementary lymphadenectomy in the case of micrometastasis, the false negative rate would have been 18%. Conclusions: After previous surgical resection of early breast cancer, ASLND remains feasible with a low identification rate of 86%, despite the use of the combined method. The False negative rate is acceptable with the use of IHC. Citation Information: Cancer Res 2012;72(24 Suppl):Abstract nr P1-01-21.