The aim of this study was to compare the incidence of lymphoedema after different treatments of the axilla in patients with breast cancer (BC). Medical records of 205 women (median age 61 years, range 26-72 years) who underwent curative surgery for primary BC were reviewed. According to the treatment of the axilla, the study population was divided into four age- and stage-matched groups of patients: Group A (N=54 patients), sentinel node biopsy (SLNB) alone; Group B (N=48 patients), SLNB followed by axillary node (AN) sampling using ultrasound scissors (harmonic scalpel); Group C (N=53 patients), AN dissection using ultrasound scissors; Group D (N=50 patients), traditional AN dissection. The median follow-up was 22 months (range 18-28 months). The intraoperative frozen section of SLNB (Groups A and B) showed 32 out of 102 (31.4%) patients with metastasis to AN, while final pathology showed AN metastases in 20, 17, 16 and 17 patients of groups A, B, C and D, respectively (p=NS). The sensitivity of SLNB alone was 80% and that of SLNB followed by AN sampling was 95% (p=NS). At follow-up patients with lymphoedema were 2 (3.7%), 2 (4.2%), 3 (5.6%) and 8 (16%) in groups A, B, C and D, respectively (p=NS). In conclusion, AN sampling is a sensitive and low-morbidity procedure which, in conjunction with the use of harmonic scalpel, may reduce the onset of arm lymphoedema.