Particulate matter air pollution and diesel engine exhaust have been classified as carcinogenic for lung cancer, yet few studies have explored associations with liver cancer. We used six European adult cohorts which were recruited between 1985 and 2005, pooled within the Effects of low-level air pollution: A study in Europe (ELAPSE) project, and followed for the incidence of liver cancer until 2011 to 2015. The annual average exposure to nitrogen dioxide (NO2), particulate matter with diameter <2.5 mu m (PM2.5), black carbon (BC), warm-season ozone (O-3), and eight elemental components of PM2.5 (copper, iron, zinc, sulfur, nickel, vanadium, silicon, and potassium) were estimated by European-wide hybrid land-use regression models at participants' residential addresses. We analyzed the association between air pollution and liver cancer incidence by Cox proportional hazards models adjusting for potential confounders. Of 330 064 cancer-free adults at baseline, 512 developed liver cancer during a mean follow-up of 18.1 years. We observed positive linear associations between NO2 (hazard ratio, 95% confidence interval: 1.17, 1.02-1.35 per 10 mu g/m(3)), PM2.5 (1.12, 0.92-1.36 per 5 mu g/m(3)), and BC (1.15, 1.00-1.33 per 0.5 10(-5)/m) and liver cancer incidence. Associations with NO2 and BC persisted in two-pollutant models with PM2.5. Most components of PM2.5 were associated with the risk of liver cancer, with the strongest associations for sulfur and vanadium, which were robust to adjustment for PM2.5 or NO2. Our study suggests that ambient air pollution may increase the risk of liver cancer, even at concentrations below current EU standards.