Introduction: Surgery could be indicated in selected patients with breast cancer liver metastases (BCLM) but accurate identification of candidates for surgery remains a challenge. We investigated if an observational period of time after the diagnosis of BCLM could improve the selection for surgery. Methods: We performed a retrospective analysis of 72 consecutive patients operated for BCLM between 2000 and 2015. Clinicopathologic and outcome data were compared between 2 groups: Early surgery (ES), including patients operated for BCLM ≤12 months after liver metastases (LM) diagnosis, and late surgery (LS), including patients operated >12 months after diagnosis. Results: Mean age was 48 years (28-74) at time of liver surgery. Median time between primary and LM diagnosis was 35 months (0-211) and median time between LM diagnosis and surgery was 12 months (0-134). Clinicopathologic characteristics were similar in ES (n=37) and LS (n=35) patients. In the whole population, 1, 3 and 5-years progression free (PFS) and overall (OS) survivals were respectively of 70%, 43%, 30% and 93%, 66%, 43%. In multivariate analysis, neither PFS nor OS differences were observed between ES and LS groups (PFS - HR=1,13 p=0,72 and OS - HR=1,06, p=0,85 for LS), only primary tumor estrogen receptor positivity had a positive impact for OS (HR=0,48 p=0,05). Conclusion: No difference in survival was observed between patients operated rapidly for BCLM as compared with patients operated after observational period, suggesting that it is not necessary to postpone the surgery in patients with resectable LM.