57.8% (26/45) of breast cancer tissues, in 20.0% (2/10) of benign tumor tissues, and in 10.0% (1/10) of noncancerous breast tissues (Chi-square test, p<0.05). And it was detected in 16.7% (2/12) of metastatic lymph nodes. 3) The positivity of EGFR of stomach cancers was gradually increased from 25.0% of stage 1, 45.8% of stage 2, 61.8% of stage 3 to 77.8% of stage 4 (Bartholomew test, p<0.01). Moreover the positivity was gradually increased according to tumor invasion and lymph node metastasis (relationship of expression of EGFR with histological type, distant metastasis, localization of cancer. The positivity of EGFR of human colon cancers was also gradually increased from 12.5% of stage A, 22.2% of stage B1, 16.7% of stage B2, 42.8% of stage C1 to 76.9% of stage C2 (Bartholomew test, p<0.05). And the positivity of EGFR was related to their depth of invasion of cancer cells (Chi-square test, p<0.05). There was no relation between the positivity of EGFR and the size, site of cancer. The EGFR was expressed only in well differentiated cancers (47.6%, 10/21) and moderately differentiated (34.8%, 8/23) cancers, while not expressed in poorly differentiated (0.0%, 0/5) colon cancers (Bartholomew test, p<0.05). The positivity of EGFR of breast cancers was also related to clinical stage; positivity of stage 1 was 60.0%, stage 2 was 62.1%, and stage 3 was 45.5%, respectively (Chi-square test, p<0.05). The positivity of EGFR was related to breast tumor size; positivity of tumor size<2㎝ was 60.0%, tumor size 2-5㎝ was 61.3%, tumor size>5㎝ was 44.4%, respectively (Chi -square test, p<0.05). 4) There was no correlation of the staining percentage of EGFR with clinical stage, tumor invasion, lymph node metastasis, and distant metastasis of stomach cancers. The staining intensity was in inverse proportion to clinical stage and lymph node metastasis (measure of association, gamma, p<0.005, p<0.005, respectively), while no correlation was observed with distant metastasis. The staining intensity of EGFR was gradually increased according to clinical stages of colon cancers (measure of association, gamma, p<0.05). It was increased in stage 1 and 2 than stage 3 of breast cancers (measure of association, gamma, p<0.05). 5) The patients with EGFR-positive adenocarcinoma had much worse prognosis than those with EGFR-negative adenocarcinoma (log-rank test, p<0.0005). The mean survival duration of the patients with EGFR-positive was shorter than those with EGFR-negative (32 months vs. 48 months, median survival). The expression of EGFR had an inverse relation between both estrogen and progesterone receptor expressions (Chi-square test, p<0.05). Conclusion: These results suggested that this improved immunohistochemical technique offers a rapid, economic, reliable and sensitive method for detection of EGFR. In stomach cancer, the expression of EGFR may act on the tumor invasiveness and lymph node metastasis to some degree, but not on distant metastasis. In colon cancer, the more the clinical stage progress, the more it may act on tumor invasiveness and metastasis. In breast cancer, it may play an important role in early stage than late. Namely, the expression of EGFR may act on different stage of each cancer and may serve as a useful prognostic marker of malignancy in patients with stomach, colon, and breast cancer in Korea.