Few studies have attempted to critically identify patient- and tumor-related factors that limit sentinel node biopsy (SNB). These studies have been limited by sample size and surgeon variability. The present study attempts to enumerate these limitations in a unique group of patients. One hundred twenty-five SNBs performed by a single surgeon between May 1997 and June 2001 were reviewed. Overall SNB was successful in 96 per cent of patients with a 97 per cent correlation with the axillary node dissection. Sentinel node identification was not affected by age, tumor size, tumor location, prior segmental resection, or neoadjuvant therapy. No false negatives were noted in the neoadjuvant group. The use of blue dye alone significantly understaged patients when compared with isotope alone (P = 0.02). SNB is a highly accurate method to identify axillary metastases and its limitations are not affected by patient or tumor related factors. In the present study SNB detection by both isotope and blue dye has been shown to be superior to blue dye alone. This finding demonstrates that these limitations may be overcome with the standardization of the technique used.