The KRAS protein is known to play a key role in various oncogenic pathways. KRAS mutations are found in 20-30 % of patients with non-small cell lung cancer (NSCLC). The majority of mutations are found at KRAS codons 12 and 13, and they appear to be more frequent in smokers and adenocarcinoma. The mutated protein is in its active state despite absence of stimulation, which leads to the constitutive activation of downstream pathways responsible for cellular disorder. The identification of new biomarkers to predict the evolution of cancer pathogenesis (predictive factor) and the sensibility to treatments (predictive factor) is one of the major objectives in oncology research. KRAS mutations are a potential candidate biomarker and numerous studies have tried to confirm its place as a prognostic/predictive biomarker in NSCLC. The results are contradictory and most studies are retrospective. The first results of prospective studies are currently reported, in particular with the use of antibodies against EGFR. The exact place of KRAS in medical thoracic oncology remains to be determined and further studies are needed. To date, KRAS mutations are not a biomarker to be used routinely.