OBJECTIVE To retrospectively assess the efficacy of post-mastectomy re-irradiation for local relapse of breast cancer. METHODS Twenty patients, initially treated by conservative surgery and radiotherapy (50 Gy in 25 fractions over 5 weeks) were treated from 1998 to 2001 for a local relapse by salvage mastectomy and re-irradiation (either electron or photon beams). Mean age was 53 years (31-71). Reasons for re-irradiation were that the local relapses were inflammatory (4 pts), multifocal (5 pts), cutaneous (5 pts), involved the nipple (3 pts) or because the surgical margins (either muscle or skin) were involved (3 pts). The median dose of re-irradiation was 45 Gy (33-65) in 15 fractions over 33 days. Mean follow-up was 48 months (5-97). RESULTS Fifteen patients remained free of a second local recurrence and 10 were still alive, without metastasis. Neither the dose of re-irradiation nor the irradiated surfaces were prognostic factors of local control (P = 0.877 and P = 0.424). Five patients developed radiation-induced pneumonitis without functional respiratory impairment. The incidence of pneumonitis seemed to be related to the biological dose of re-irradiation (P = 0.037). Other late complications occurred such as pigmentation changes (12 pts), telangiectasia (8 pts), chondritis (2 pts), parietal fibrosis (7 pts), rib fractures (4 pts), severe pain (11 pts) and lymphedema (2 pts). The increase in biological equivalent dose was highly statistically linked with the occurrence of disabling pain (P = 0.0123). CONCLUSIONS Parietal re-irradiation achieves good and lasting local control with an acceptable rate of acute complications but with a risk of disabling late sequelae such as severe pain.