The purpose of this study was to analyze whether inter-site variation types on estrogen receptor (ER) and HER2 expression may be a predictive factor for evaluating the effectiveness of endocrine therapy in patients with ER-positive and HER2-positive breast cancer. A total of 366 consecutive women with invasive breast cancer who had undergone curative surgical treatment between 1996 and 2001 were included in this study. ER status was evaluated using the Allred score and HER-2 status was evaluated according to the HercepTest. In ER-positive and HER2-positive tumors, the expression of ER and HER2 was described as the co-expressed type or the differently expressed type using double staining with ER and HER2. Of the 366 patients, 249 (68.1%) were positive for ER and 74 (20.2%) were positive for HER2. ER-positive and HER2-negative tumors were found in 221 patients (60.4%), ER-negative and HER2-negative in 71 (19.4%), ER-negative and HER-2-positive in 46 (12.6%), and ER-positive and HER2-positive in 28 (7.7%). HER2 status was inversely correlated (p<0.01) with ER status. In ER-positive tumors, an inverse correlation between ER and HER2 was also observed. The co-expressed type was found in 10 patients, and the differently expressed type was found in 18. There was no difference in tumor size and nodal involvement between the two types. There was no significant difference in disease-free survival between patients with the co-expressed type tumor and the differently expressed type tumor. In patients with the differently expressed type tumor, those who received antiestrogen therapy showed a significantly better disease-free survival rate than those who did not receive antiestrogen therapy. As for patients with the co-expressed type of tumor, no significant difference in disease-free survival was observed between patients with and without antiestrogen treatment. The present study suggests that the co-expressed type of tumour might be a resistant factor to antiestrogen therapy in ER-positive and HER2-positive breast cancer.