An epidermal growth factor receptor (EGFR) has been reported to be associated with a poor clinical outcome in breast cancer, while its prognostic value remains controversial. Immunohistochemical staining for EGFR was performed on frozen sections of primary breast cancer from 1029 patients with a mean follow-up duration of 46 months. EGFR was positive in 277 (26.9%) of 1029 cases which inversely correlated with the estrogen receptor (ER) status. A univariated analysis indicated that EGFR had a significant prognostic value in both the disease free survival (DFS) and the overall survival (OS), while the same effect was also found in node negative as well as node positive breast cancer. A multivariate analysis indicated that EGFR was an independently significant prognostic factor for DFS (p = 0.0174) and OS (p = 0.0105) in all patients, but that EGFR demonstrated a prognostic significance only for DFS (p = 0.0241) in node negative and only for OS (p = 0.0333) in node positive breast cancer. When all patients were stratified for EGFR and ER, a multivariate analysis indicated that the combination of EGFR(+)/ER(-) was an independently significant factor for both DFS and OS in node negative as well as node positive breast cancer. In conclusion, the prognostic value of EGFR was demonstrated by a multivariate analysis in a large series of breast cancer patients, but the value of EGFR was somewhat insufficient to achieve statistical significance for both DFS and OS in the subgroups divided by nodal status. On the other hand, the prognostic value of combination of EGFR and ER was sufficient to achieve statistical significance based on a multivariate analysis for both DFS and OS in the subgroups of node negative as well as node positive breast cancer patients.