Lung cancer is an important public health problem with almost no improvement in survival over the past decades. Although observational studies demonstrate that low-dose multi-detector spiral-CT screening is able to detect lung cancer in an early stage in 55-85% of all cancer cases detected, and that 5- and even 10-year survival rates close to 90% can be achieved, these studies do not answer the question whether CT screening is advisable. Excellent survival rates in a few individuals do not necessarily indicate that there is a lung cancer-specific reduction in mortality, since observational studies are subject to several biases: lead time bias, over-diagnosis bias, and length time bias. Therefore, there is a strong worldwide recommendation from various professional organisations not to adopt CT screening for lung cancer on a wide scale, but to await the results from large randomised studies such as the US 'National lung screening trial' and the Dutch-Belgian-Danish 'Netherlands-Louvain lung cancer screening study' (NELSON), which will provide more clarity as to the effectiveness and cost-effectiveness and possible negative effects of CT screening for lung cancer.