Clinical and pathologic staging of lung cancer is suboptimal in achieving the goals of assessing prognosis and selecting therapy. Although the technologic developments that allow the generalized use of proteomic and genomic analyses are relatively recent, major progress in understanding the molecular basis of lung cancer has been made. Predicting survival is only the first step in the use of genomics and proteomics. If a reliable gene array or protein profile can be identified that is associated with poor prognosis, these profiles can then be identified and become potential therapeutic targets. It is not difficult to envision the development of a simple serum test that will diagnose a lung cancer perhaps even before it is clinically apparent and at the same time identify the chemotherapeutic agents to which the tumor is sensitive, allowing individually directed treatment. Eventually, a comprehensive staging system should incorporate the prognostic information of biologic variables.