To estimate the relationship between air pollution and lung cancer, a nested case-control study was set up within EPIC (European Prospective Investigation on Cancer and Nutrition). Cases had newly diagnosed lung cancer, accrued after a median follow-up of 7 years among the EPIC ex-smokers (since at least 10 years) and never smokers. Three controls per case were matched. Matching criteria were gender, age (+/-5 years), smoking status, country of recruitment and time elapsed between recruitment and diagnosis. We studied residence in proximity of heavy traffic roads as an indicator of exposure to air pollution. In addition, exposure to air pollutants (NO(2), PM10, SO(2)) was assessed using concentration data from monitoring stations in routine air quality monitoring networks. Cotinine was measured in plasma. We found a nonsignificant association between lung cancer and residence nearby heavy traffic roads (odds ratio = 1.46, 95% confidence interval, CI, 0.89-2.40). Exposure data for single pollutants were available for 197 cases and 556 matched controls. For NO(2) we found an odds ratio of 1.14 (95% CI, 0.78-1.67) for each increment of 10 microg/m(3), and an odds ratio of 1.30 (1.02-1.66) for concentrations greater than 30 microg/m(3). The association with NO(2) did not change after adjustment by cotinine and additional potential confounders, including occupational exposures. No clear association was found with other pollutants.