An impressive number of publications refer to prognostic and predictive factors in lung cancer. TNM classification and performance status significantly influence the choice of treatment and strongly predict patients' survival. Depending on the population studied (small cell or non-small cell cancer, operable or not) other independent factors improve the prediction of prognosis; they are clinical, biological, radiological or molecular and pertain to the tumor or the patient. Molecular targeted therapies development has renewed the interest towards predictive factors. New strategies are developed to explore individual response to treatment such as EGFR tyrosine-kinase inhibitors, without success for anti-angiogenic treatments. Conventional cytotoxic agents may also be customized with predictive factors (i.e. ERCC1 or RRM1). Large multicenter studies are needed to validate new independent prognostic factors and increase our current knowledge aiming at separating patients who will really benefit from therapies of those who will only experience the side effects.