Inhalatory intake of environmental agents may have adverse effects on health, the lung being the first target. Therefore, an increased risk of lung cancer and respiratory disease is in general considered as an indication of environmental health problems related to exposure to industrial emissions, traffic exhaust and smog. Classical epidemiological studies of the association between exposure to ambient air pollutants and respiratory dysfunctions and studies with laboratory animals have failed to demonstrate the distinct proof of risk for the general population that would be needed to form a basis for high impact environmental policy measures. Here, as an example, we describe the uncertainty in assessing risks of lung cancer associated with environmental exposure to polycyclic aromatic hydrocarbons. The recently introduced methodology of molecular cancer epidemiology is considered to yield more information on the relationship between exposure to environmental carcinogens and tumour development. Recent advances in the study of carcinogen (polycyclic aromatic hydrocarbon) dosimetry at the DNA level in combination with proto-oncogenic activation in humans are described.