To evaluate the significance of the pathology margins of the tumor excision on the outcome of treatment, an analysis was performed of 697 consecutive women with clinical Stage I or II invasive carcinoma of the breast treated with breast-conserving surgery and definitive irradiation. Complete gross excision of the primary tumor was performed in all cases, and an axillary staging procedure was performed to determine pathologic axillary lymph node status. The 697 patients were divided into four groups based on the final pathology margin from the primary tumor excision or from the re-excision if performed. These four groups were: (a) 257 patients with a negative margin (greater than 2 mm), (b) 57 patients with a positive margin, (c) 37 patients with a close margin (less than or equal to 2 mm), and (d) 346 patients with an unknown margin. The patients with positive final pathology margins were focally positive on microscopic examination. Patients with grossly positive margins or with diffusely positive microscopic margins were treated with conversion to mastectomy. There was a significant difference in the total radiation dose for the four groups (median dose of 6000 vs 6500 vs 6400 vs 6240 cGy, respectively; p less than .0001). There was no significant difference among the four groups for 5-year actuarial overall survival (p = .19), no evidence of disease (NED) survival (p = .95), or relapse-free survival (p = .80). There was no significant difference among the four groups for five year actuarial local or regional control (all p greater than or equal to .29). Subset analyses did not identify any poor outcome subgroups. These results have demonstrated that selected patients with focally positive or close microscopic pathology margins can be adequately treated with definitive breast irradiation. Patient selection and the technical delivery of radiation treatment including a boost may have been important contributing factors to the good outcome in these patients.