OBJECTIVE: Although large-scale studies, including meta-analyses, on the association between alcohol consumption and lung cancer have been reported, the risk of lung cancer in extremely high consumption groups (>or=450 g ethanol/week) in both current smokers and nonsmokers remains unclear. Here, we investigated the association between alcohol consumption and lung cancer incidence. METHODS: We conducted a population-based prospective cohort study in 46,347 Japanese men aged 40-69 years with no history of cancer at baseline in 1990-1994. We used Cox proportional hazards regression analysis to estimate the hazard ratio (HR) and 95% confidence interval (CI) of lung cancer incidence according to alcohol consumption. RESULTS: During the 14-year follow-up period, we documented 651 cases of lung cancer. We found a positive association between alcohol consumption and lung cancer risk confined to current smokers. Compared with occasional drinkers, the multivariate HRs (95% CI) in the highest category (>or=450 g ethanol/week) were 1.31 (0.89-1.94) among total participants, 0.58 (0.26-1.30) (p for trend = 0.49) among nonsmokers, and 1.69 (1.05-2.72) (p for trend = 0.02) among current smokers. CONCLUSIONS: Among this population with a large variation in alcohol consumption, alcohol consumption was not an independent risk factor for lung cancer except for current smokers.