Aberrant expression of the epidermal growth factor receptor (EGFR) system has been reported in a wide range of epithelial cancers. In some studies, this has also been associated with a poor prognosis and resistance to the conventional forms of therapies. These discoveries have led to the strategic development of several kinds of EGFR inhibitors, five of which have gained US Food and Drug Administration approval for the treatment of patients with non-small-cell lung cancer (gefitinib and erlotinib), metastatic colorectal cancer (cetuximab and panitumumab), head and neck (cetuximab), pancreatic cancer (erlotinib) and breast (lapatinib) cancer. Despite these advances and recent studies on the predictive value of activating EGFR mutation and KRAS mutations with response in non-small-cell lung cancer and colon cancer patients, there is currently no reliable predictive marker for response to therapy with the anti-EGFR monoclonal antibodies cetuximab and panitumumab or the small molecule EGFR tyrosine kinase inhibitors gefitinib and erlotinib. In particular, there has been no clear association between the expression of EGFR, determined by the US Food and Drug Administration-approved EGFR PharmDX kit, and response to the EGFR inhibitors. Here, we discuss some of the controversial data and explanatory factors as well as future studies for the establishment of more reliable markers for response to therapy with EGFR inhibitors. Such investigations should lead to the selection of a more specific subpopulation of cancer patients who benefit from therapy with EGFR inhibitors, but equally to spare those who will receive no benefit or a detrimental effect from such biological agents.