In a study on 214 patients with primary breast cancer (median follow-up 8.5 yr, maximum follow-up 15 yr), EGF-R was negatively correlated to estrogen receptor and progesterone receptor, whereas no association was found with age, lymph node status, and tumor size. Initially, after a follow-up of 5 yr, there was a tendency to a significant association between EGF-R levels and tumor recurrence rate (p = 0.08). Patients with tumors containing intermediate levels of EGF-R experienced a longer relapse-free survival (RFS) than did patients with tumors possessing lower or higher levels of EGF-R. This effect was most pronounced in the subgroup of patients with positive axillary lymph nodes. However, after 10 yr follow-up, this association appears to be lost (p = 0.28) as shown in this update. A similar phenomenon was observed for the ER. While at 5 yr follow-up ER status had significant prognostic value (p = 0.01), at 10 yr follow-up this significance also appears to be lost (p = 0.40). However, tumor size, lymph node status, grade, and PgR status maintained significant prognostic value by univariate analysis. Based on 40 separate studies comprising 5232 patients, the mean percentage of EGF-R positivity reported in breast cancer is 45% (range 14-91%). Nine out of 15 different studies showed in some way a significant negative association between EGF-R and RFS by univariate analysis, and 2 others showed a tendency to such a relationship. Of 7 studies applying multivariate analysis, two demonstrated an independent prognostic value of EGF-R for RFS and two others a tendency to a significant correlation, whereas three did not. It may be concluded that EGF-R status has more or less prognostic value in patients with primary breast cancer, but the prognostic power decreases with longer follow-up. Of great clinical significance is the association of EGF-R with hormone resistance. Therefore EGF-R status can be used for selection of type of treatment. Finally, EGF-R might be useful as a target for new treatment modalities.