The EGF receptor (EGFR) has emerged as a rational target for anticancer therapy for the treatment of colorectal cancer (CRC). Positive immunohistochemistry (IHC) staining for EGFR is used as a criterion for patient selection; however, doubt has been cast on the utility of this method. Not only is the response to cetuximab, an anti-EGFR mAb, low in patients expressing EGFRs, but a similar response to cetuximab has also been described in patients who do not express EGFRs. This review aims to evaluate the possible cause of the lack of correlation between the efficacy of cetuximab and EGFR IHC staining in CRC, as well as any modifications in the IHC method necessary to optimize patient selection for cetuximab therapy. In our opinion, the heterogeneous expression of the receptor in the neoplastic population and the inability of the mAbs used to predict the response to cetuximab could be the major cause of the failure of IHC staining as a reliable tool for patient selection. The use of specific mAbs directed against the phosphorylated and mutant form (EGFRvIII) of the EGFR could reinstate IHC as a valid predictor of response to therapy.