Non-small cell lung cancer (NSCLC) is a heterogeneous disease, caused by the presence of different clinically relevant molecular subtypes. Genetic mutations are emerging as potential biomarkers of response and treatment selection in patients with NSCLC. Over the past few years, activating mutations of epidermal growth factor receptor (EGFR) have been recognized as the most important predictor of response to EGFR tyrosine kinase inhibitors (TKIs) such as gefitinib and erlotinib and also as a favourable prognostic factor. The RAS genes, including H-RAS, K-RAS and N-RAS, encode a family of proteins regulating cell growth, differentiation and apoptosis. Mutations in the K-RAS gene, mainly in codons 12 and 13, have been found in 20-30% of NSCLC tumor samples and occur most commonly, but not exclusively, in adenocarcinoma histology and in heavy smokers. In NSCLC, the presence of K-RAS mutations has generally been considered to be associated with worse prognosis and resistance to systemic therapy in the adjuvant as well as the metastatic setting. In early stage NSCLC, the prognostic role of K-RAS mutations has been evaluated in several studies without definitive conclusion. On the other hand, in advanced NSCLC, the presence of K-RAS mutations identifies a subgroup of patients who do not respond to EGFR-TKI therapy but, at the same time, a positive survival effect from EGFR-TKIs cannot be excluded in these patients. Similarly, K-RAS mutational status does not predict benefit from the anti-EGFR monoclonal antibody cetuximab, highlighting the biological difference between lung cancer and colorectal cancer. As a result of the lack of conclusive data, K-RAS mutations do not represent a validated biomarker for the negative selection of patients who are candidates for anti-EGFR therapy. The aim of this article is to review and discuss the data on the prognostic and predictive value of K-RAS mutations in NSCLC.