Lung cancer incidence in general middle-aged men estimated by low-dose computed tomography screening has been unknown. We conducted a longitudinal study to estimate the incidence of lung cancer in general middle-aged Japanese men based on smoking status and to determine the optimal targets for lung cancer screening. The subjects were 14,058 Japanese men aged 40-59 years (mean age, 48.3 years; median age, 49 years) who underwent twice or more low-dose computed tomography screenings at Nippon Telegraph and Telephone East Corporation Izu Hospital at the baseline examination during 2000-2007. Medical records of subjects with abnormal findings were collected and analyzed. The lung cancer incidence in men aged 40-59 or 50-59 years based on smoking status by person-years method, and relative risks of current- and former-smokers compared with non-smokers were calculated. Mean follow-up period was approximately 3 years, and total person-years were 45,152. The incidence was 24.4 per 100,000 person-years in men aged 40-59 years (95% confidence interval (95% CI) 13.6-43.6), and 56.5 in men aged 50-59 (95% CI 31.6-101). No statistically significances were observed among non-, current and former-smokers. However, the relative risk of the current-smokers aged 50-59 years with 40 or more pack-years, compared with non-smokers was 6.31 (95% CI 1.27-31.3), and was statistically significant. No significant differences in the lung cancer incidence existed among any smoking status in the Japanese men aged 40-59 years. In this population, men aged 50-59 years with 40 and more pack-years could be an optimal target for lung cancer screening. The small number of incident lung cancers resulting in the incidence and relative risks with the wide 95% CI and the participants' demographic biases were the limitations of our study. Our screening recommendations should be made with more caution. (C) 2008 Elsevier Ireland Ltd. All rights reserved.