The purpose of this article was to review the patterns and morbidity of regional recurrence (RR) in patients with early breast cancer, efficacy of salvage therapy for RR, and complications of regional nodal treatment. A retrospective evaluation of 1,158 patients with stage I or stage II breast cancer treated with conservative surgery and radiotherapy (RT) between 1979 and 1994 was performed. Seven hundred fifty patients underwent axillary surgery, and 229 patients received RT as their only treatment of the regional lymphatics. Regional nodal RT was given to 168 patients who also had axillary surgery. The regional lymphatics of 11 patients were not treated. The patterns and morbidity of RR, relapse management, and complications related to regional nodal treatment were reviewed from the patients' records. With a median follow-up of 88 months, a total of 31 patients (2.7%) developed a RR. Nine of 31 patients (29%) with an RR experienced significant morbidity, including pain, fungating tumor, dysphagia, dyspnoea, and/or sensory motor changes at diagnosis. Nineteen patients (61%) had symptomatic residual or progressive regional disease after salvage therapy at last follow-up or death. Six of nine patients (67%) who developed an isolated axillary recurrence and underwent salvage surgery had no further axillary recurrence. The addition of regional nodal RT to breast irradiation significantly increased the incidence of symptomatic pneumonitis (1% without regional nodal RT and 4% with regional nodal RT, p < 0.001). Combined axillary dissection and nodal irradiation resulted in a significantly higher incidence of arm edema compared with either alone (9.5% with axillary dissection, 6.1% with RT to the axilla and supraclavicular fossa, and 31% with combined modality therapy, p < 0.001). Five of 380 patients (1%) who received RT to the axilla and/or supraclavicular fossa developed a transient brachial plexus neuropathy. Although RR was uncommon in patients treated with axillary surgery and/or regional nodal irradiation, salvage therapy failed to eradicate the recurrence in approximately two thirds of the patients with a RR. Ongoing research is essential to optimize regional control with an acceptable level of risk of treatment complications. Sentinel lymph node biopsy, if validated as an accurate method of staging the axilla in patients with breast cancer, would allow selective avoidance of regional nodal treatment and hence the associated morbidity.