Breast-conserving surgery (BCS) is the treatment of choice for women with early-stage BC. Several studies indicated a better outcome in patients undergoing BCS plus radiotherapy (RT) in comparison with mastectomy (MAST) +/- RT (PMID:33950173, 27344114). Indeed, in comparison with MAST, BCS reduces the extent of surgery, but the addition of radiotherapy (RT) makes it hard to understand whether the differences in outcome may be attributed to the extent of the surgery only. We performed a competing risks re-analysis in terms of dynamics and crude cumulative incidence (CCI), of distant recurrence free survival (DRFS) from two historical randomized clinical trials which of the Istituto Nazionale dei Tumori of Milan: the “Milan 1” trial (n = 701; 1973-1980, PMID:7015141) which compared MAST with BCS plus RT (BCS+RT) and the “Milan 3” trial (n = 567; 1987-1989, PMID: 8387637) which adopted the same surgical conservative approach with (BCS +RT) or without (BCS -RT) RT. Clinical features such as primary tumour size, axillary lymph node status (N) and menopausal status were considered. Concerning distant recurrence (DR) dynamics in Milan 1 trial, evidence of a different intensity of cause-specific hazard was found among subgroups related to surgery and lymph-node status (test of interaction effect: P=0.005), though with similar multi-peaked hazard patterns. Indeed, we observed that a MAST-related worse outcome is present in women with positive N (N+). Analysis of CCI reveals that, for the subgroup of N+ patients in Milan 3 trial, removal of RT from BCS (BCS -RT) results in 30% higher incidence of DR, comparable to the difference between MAST N+ and BCS+RT N+ patients in the Milan 1 trial. In this re-analysis of historical randomized clinical trials, we report that the worse outcome of MAST in comparison with BCS may be mainly attributed to higher extensive surgery and that the effect of the extent of the surgery without RT is mostly confined to patients with N+ BC. The effect of RT, as assessed by the re-analysis of Milan 3 trial, is likely related to a combined effect of the control of local recurrences, and the potential systemic effect of RT.