Exposure to metallic compounds is ubiquitous, with its widespread use in industry and its presence, mostly in trace amounts, in the environment. This paper reviews the epidemiologic evidence of the relation between lung cancer and exposure to metallic compounds by building on and updating the corresponding International Agency for Research on Cancer (IARC) assessments. Given that most of the well-identified human populations with given metal exposure are in occupational settings, this review is mostly based on results in occupational epidemiology. The epidemiological evidence is shortly reviewed for accepted carcinogens: chromium, nickel, beryllium, cadmium, arsenic, and silicon, highlighting what is still unclear. We then review in more detail metals for which the evidence is less clear: lead, titanium, iron, and cobalt. There is scarce evidence for the human carcinogenicity of titanium. Exposure to titanium dioxide is associated with lung cancer excesses in one large study, but this excess may be due to confounders. The evidence for lead is contradictory. The lung cancer risk is presented as a function of a post hoc exposure ranking but no dose-response relationship is found. A weak but consistent lung cancer excess in many populations exposed to iron oxides but it is not possible to state on causality. Finally the evidence in the hard metal industry is presented, which suggests a possible carcinogenic effect of cobalt in presence of tungsten carbide. A short discussion presents the limitations of epidemiology in assessing the carcinogenicity of metals.