Molecules interfering with the Epidermal Growth Factor Receptor (EGFR) have been successfully tested in several human malignancies in the last decade, including non-small cell lung cancer, colo-rectal, pancreatic and head and neck cancer. Particularly, the two most commonly used strategies for blocking EGFR include tyrosine-kinase inhibitors (TKIs) targeting the intracellular domain of the receptor and monoclonal antibodies (MAb) directed against its external portion. One of main goals of researchers is to identify biological predictors of activity or resistance to these agents, both for ethical and pharmacoeconomical reasons. EGFR protein expression assessed by immunohistochemistry does not seem to accurately predict activity of either class of compounds, while presence of EGFR sensitizing mutations, which can be found in significant fractions of NSCLC patients, has been associated with a better outcome in patients receiving EGFR TKIs. Increased EGFR gene copy number could represent a reliable and reproducible tool for proper selection of patients candidate to TKIs or anti-EGFR MAbs. Furthermore, mechanisms of resistance to anti-EGFR strategies are currently being elucidated, allowing identification of subjects who should be excluded from treatment.