INTRODUCTION: Recent and historical trials have suggested that the omission of axillary surgery is oncologically safe in node-negative early breast cancer. This meta-analysis investigates the feasibility of the omission of axillary surgery (SLNB or ALND) in terms of oncological outcomes and adjuvant treatment decisions. METHOD: A systematic search of Medline, Embase and Cochrane Central was conducted. Random-effect meta-analysis was conducted to compare recurrence and survival outcomes between omission of axillary surgery and SLNB or ALND. Differences in real-world adjuvant treatment decisions and patient reported outcomes were investigated. RESULTS: Ten studies (omission, n = 3716; SLNB/ALND, n = 4604/785) investigated oncological outcomes (pooled rates, T1 88 %; Grade 1-2 80 %; ER-positive 91.2 %; HER2-positive 4.4 %; ductal carcinoma 74 %). In the omitted group, the pooled rates of local, axillary and distant recurrence (follow-up, 8 years 9 months) were 3.0 % (95 %CI,1.5-5.9 %), 2.5 % (95 %CI,1.3-4.8 %; 5year, 1.0 %) and 3.6 % (95 %CI,2.0-6.5 %; 5year, 2.7 %), respectively. When comparing omission to SLNB/ALND, there were no differences in local recurrence (OR 0.91; 95 %CI,0.56-1.50), distant metastasis (OR 0.91; 95 %CI,0.56-1.50), BCM (OR 1.00; 95 %CI,0.63-1.60), OS (HR 0.88; 95 %CI,0.65-1.19) or DFS (HR 0.91; 95 %CI,0.76-1.11). Axillary recurrence was higher in the omission group compared to SLNB/ALND (OR 4.42; 95 %CI,1.5-12.8) but not in SLNB alone (OR 2.85; 95 %CI,0.1-133.8). Omission of axillary surgery was associated with lower rates of adjuvant chemotherapy, radiotherapy and hormonal treatment but quality of evidence was poor. CONCLUSION: Omission of axillary surgery is safe in node-negative early breast cancer. Prospective data is required to investigate the impact of omission on adjuvant treatment decision.