BACKGROUND: Particulate matter (PM) air pollution is a human lung carcinogen; however, the components responsible have not been identified. We assessed the associations between PM components and lung cancer incidence. METHODS: We used data from 14 cohort studies in eight European countries. We geocoded baseline addresses and assessed air pollution with land-use regression models for eight elements (Cu, Fe, K, Ni, S, Si, V and Zn) in size fractions of PM2.5 and PM10. We used Cox regression models with adjustment for potential confounders for cohort-specific analyses and random effect models for meta-analysis. RESULTS: The 245,782 cohort members contributed 3,229,220person-years at risk. During follow-up (mean, 13.1years), 1878 incident cases of lung cancer were diagnosed. In the meta-analyses, elevated hazard ratios (HRs) for lung cancer were associated with all elements except V; none was statistically significant. In analyses restricted to participants who did not change residence during follow-up, statistically significant associations were found for PM2.5 Cu (HR, 1.25; 95% CI, 1.01-1.53 per 5ng/m(3)), PM10 Zn (1.28; 1.02-1.59 per 20ng/m(3)), PM10 S (1.58; 1.03-2.44 per 200ng/m(3)), PM10 Ni (1.59; 1.12-2.26 per 2ng/m(3)) and PM10 K (1.17; 1.02-1.33 per 100ng/m(3)). In two-pollutant models, associations between PM10 and PM2.5 and lung cancer were largely explained by PM2.5 S. CONCLUSIONS: This study indicates that the association between PM in air pollution and lung cancer can be attributed to various PM components and sources. PM containing S and Ni might be particularly important.