The hypotheses that the antioxidant vitamin E provides protection against lung cancer and that this hypothetical protection is modified by smoking status were investigated using two different study designs--a cohort study and a nested case-control study--among Finnish men aged 15 years and over. In the cohort study the association between vitamin E intake and lung cancer risk was studied among 5,254 individuals with 121 lung cancer cases that occurred during a 19-year follow-up, and in the nested case-control study the association between serum vitamin E level and lung cancer risk was studied using 144 lung cancer cases and 270 matched controls as a basis. There was a significant inverse association between vitamin E status and lung cancer occurrence among nonsmokers but not among smokers in both designs. The relative risk of lung cancer between the lowest and highest tertiles of vitamin E intake was 3.3 among nonsmokers and 0.8 among smokers. The corresponding results for serum vitamin E were 6.6 and 0.8, respectively. Nonsmokers with simultaneously low serum levels of vitamin E and other micronutrients (i.e., beta-carotene, retinol and selenium) had a 12-fold greater risk of lung cancer in comparison with men having more satisfactory levels. The corresponding number among smokers was three. The results suggest that vitamin E status is primarily associated with lung cancer risk among nonsmokers. Firm conclusions can, however, be drawn only on the basis of intervention trials.