Background/Objective Cavity shave margins (CSMs) decrease rate of positive margins and need for re-excision. Recurrence data following breast-conserving surgery (BCS) are not always available in large cancer registries. We sought to define our recurrence and survival data in BCS with routine excision of CSMs. Methods A single institution, 10-year retrospective review of breast cancer patients who underwent BCS with routine CSMs was conducted. Cavity shave margin technique was standard. Cox proportional hazard analyses and the Kaplan-Meier method were used to estimate recurrence and survival. Results Breast-conserving surgery with CSM was performed in 839 patients. Re-excision rate to achieve negative margins was 8.5%. Fifty-two patients (75%) underwent margin re-excision vs 18 patients (25%) underwent salvage mastectomy. Positive margin rate stratified by tumor histology was highest for invasive lobular carcinoma followed by mixed invasive ductal carcinoma (IDC) and ductal carcinoma in situ (DCIS), followed by pure DCIS and lowest for IDC. Length of follow-up was (4.7 ± 2.6, years). Overall recurrence rate (locoregional and systemic) was 4.3%: highest in patients with negative lumpectomy margin but positive CSM (L−S+ = 15%) followed by positive lumpectomy and CSMs (L+S+ = 14%), followed by patients with positive lumpectomy margin but negative CSMs (L+S− = 13%) and lowest for negative lumpectomy and CSM (L−S− = 5%), ( P = .0008). There was no difference in 5-year breast cancer–specific survival between the 4 subgroups: 96% for L−S−, 86.7% L−S+, 94.7% L+S+ and 90% L+S− ( P = .094). Conclusions Recurrence following BCS with CSMs can be stratified based on both lumpectomy and cavity shave margin positivity. Routine excision of CSMs allows identification of these patient subsets.