This randomized controlled trial aimed to investigate the effects of circumferential shaving on reducing the intraoperative margin positivity rate (MPR) during breast-conserving surgery (BCS). Eligible breast cancer patients were randomly assigned into no-shave and shave groups. In the no-shave group, the cavity margins were collected for assessment after the tumor resection, whereas in the shave group, a circumferential shaving was performed before collecting the cavity margins. The primary outcome was the intraoperative MPR by frozen section analysis. A total of 181 patients, with a median age of 49 years, were randomized. Patient characteristics at baseline were well-balanced between the two groups. The intraoperative MPRs (12.1% vs. 7.8%, p = 0.38), postoperative MPRs (16.5% vs. 7.8%, p = 0.073), intraoperative re-excision rates (26.4% vs. 23.3%, p = 0.64), second operation rates (4.4% vs. 1.1%, p = 0.34), and successful BCS rate (93.4% vs. 94.4%, p = 0.94) were all similar between the no-shave and the shave groups. The volume of the shaved tissues was significantly increased in patients with larger breast volume (p < 0.01). In patients with C–E cup breasts, the no-shave and shave groups had 16.7% and 0% (p = 0.03) intraoperative MPRs, and 22.0% and 0% (p = 0.01) postoperative MPRs, respectively. In patients with A–B cup breasts, the MPRs were similar between the two groups. The presence of the ductal carcinoma in situ component is the only determinant of margin positivity. Circumferential shaving did not significantly reduce the MPR in BCS. Its benefit depends on the volume of the shaved tissues and the breast. Trial registration This trial was registered at ClinicalTrials.gov (NCT02648802).