BACKGROUND: Air pollution is suspected to cause lung cancer. The purpose was to investigate whether the concentration of nitrogen oxides (NOx) at the residence, used as an indicator of air pollution from traffic, is associated with risk for lung cancer. METHODS: We identified 679 lung cancer cases in the Danish Cancer Registry from the members of three prospective cohorts and selected a comparison group of 3,481 persons from the same cohorts in a case-cohort design. Residential addresses from January 1, 1971, were traced in the Central Population Registry. The NOx concentration at each address was calculated by dispersion models, and the time-weighted average concentration for all addresses was calculated for each person. We used Cox models to estimate incidence rate ratios after adjustment for smoking (status, duration, and intensity), educational level, body mass index, and alcohol consumption. RESULTS: The incidence rate ratios for lung cancer were 1.30 [95% confidence interval (95% CI), 1.07-1.57] and 1.45 (95% CI, 1.12-1.88) for NOx concentrations of 30 to 72 and >72 microg/m3, respectively, when compared with <30 microg/m3. This corresponds to a 37% (95% CI, 6-76%) increase in incidence rate ratio per 100 microg/m3 NOx. The results showed no significant heterogeneity in the incidence rate ratio for lung cancer between cohorts or between strata defined by gender, educational level, or smoking status. CONCLUSION: The study showed a modest association between air pollution from traffic and the risk for lung cancer. IMPACT: This study points at traffic as a source of carcinogenic air pollution and stresses the importance of strategies for reduction of population exposure to traffic-related air pollution.