The overall 5-year survival of lung cancer is only 10% in Europe and 15% in the United States, and progress in curative treatments during the last 20 years has been modest. Late diagnosis of extensive disease is the main reason of failure. Early detection with low-dose spiral computed tomography (CT) is one of the most promising development of clinical research, and continuous improvements in technology can make this instrument more effective than mammography in breast cancer detection. In order to prove the benefit of early detection by reduction of lung cancer mortality, we need to enrol large numbers of high-risk individuals in multicentric prospective randomized trials combining primary prevention by smoking cessation with diagnostic intervention with low-dose spiral CT, optimal management of cancer and minimum damage for healthy individuals. Molecular biology research within early detection trials, combining genomic and proteomic analysis of blood and sputum, may improve the differential diagnosis, define the individual risk of cancer incidence and failure, and help target therapies on the basis of biologic profile.