Approximately 2,800 new cases of cancer in the lungs and respiratory tract are diagnosed annually in Sweden. By the time it is diagnosed, the cancer has usually spread in the body, and the 5-year survival rate following diagnosis is approximately 10 percent. If cancer could be detected at an early stage, the hypothesis is that it would substantially improve the prognosis. Detecting lung cancer at an early, asymptomatic stage requires targeted screening. Earlier attempts to use chest radiography in early detection of lung cancer have not lowered the mortality rate from the disease. The development of spiral computed tomography (CT) and the use of lower radiation doses have, however, renewed the interest in screening. Smoking is the main cause of lung cancer. It has been estimated that it takes 20 to 30 years of smoking before lung cancer progresses to a symptomatic stage. A targeted screening program including all those who are or have been daily smokers, and are older than 45 years of age, would comprise 600,000 to 900,000 individuals annually in Sweden. Scientific studies have not confirmed that early detection of lung cancer leads to a reduction in mortality from the disease. As with assessments of screening for other types of cancer, , breast cancer, large, long-term, randomized trials would be required to answer this question. Six assessments, with major differences in design, have investigated the effectiveness of low-dose spiral CT in detecting early lung cancer. Lesions that led to follow-up examinations were detected in 5 percent to 69 percent. Between 0.4 percent and 2.7 percent of those screened were diagnosed as having lung cancer. The positive predictive value ranged between 2.8 percent and 11.6 percent. Two of the studies compared CT with chest radiography, and both studies showed that more tumors were detected with computed tomography. Most of the detected tumors were small (stage I). Detecting a single case of cancer at an early stage requires examination of a large number of people. This would suggest high costs for a screening program. Since the benefits of early detection have not been determined, model studies are the only way to estimate the cost effectiveness of the method. Screening for lung cancer requires examination of a large number of healthy individuals in order to detect disease in, and offer treatment to, a few. Therefore, one must carefully weigh the advantages and disadvantages. Beyond subjecting patients to anxiety and time sacrificed in conjunction with the examination, the use of radiation involves some risk for induction of cancer. Also, a number of patients will be subjected to unnecessary invasive procedures due to false-positive test results. Currently, there is good scientific documentation of the capacity for low-dose CT to detect lung cancer at an early stage. There is no scientific documentation to show how mortality would be affected by an organized program using low-dose CT for early detection of lung cancer. There is poor documentation on the cost effectiveness of the method. Before considering the appropriateness of routine screening for lung cancer, a positive effect on mortality must be demonstrated by large, controlled trials of screening programs.