The prognostic value of EGFR expression in colorectal cancer, usually evaluated by immunohistochemistry, is actually based on heterogeneous data, considering tumour stage or survival rates. Results are variable due to differences in evaluation criteria between studies. The development of standardized scoring systems and the evaluation of expression variability in tumour led to reconsider this question. It seems to be important to evaluate precisely EGFR reactivity in the deepest region of the tumour. The finding of a profile linked to worse prognosis could allow pointing out tumours where EGFR overexpression is associated to disease progression and therefore could help to define EGFR- inhibitors' indications in colorectal cancer.