The mortality experienced by a cohort of 10,491 United Kingdom steel foundry workers during the period 1946-85 has been investigated. These workers were all male operatives first employed in any one of the 10 participating foundries in 1946-65; all had worked in the industry for a minimum period of one year. Compared with the general population of England and Wales, statistically significant excesses relating to cancer mortality were found for cancer of the stomach (E = 77.4, O = 106, SMR = 137) and cancer of the lung (E = 229.2, O = 441, SMR = 147). A statistically significant deficit was found for cancer of the brain (E = 19.4, O = 10, SMR = 51). Involvement of occupational exposures was assessed by the method of regression models and life tables (RMLT). This method was used to compare the duration of employment in the industry, in "dust exposed" jobs, in "fume exposed" jobs, in foundry area jobs, in fettling shop jobs, and in foundry area or fettling shop jobs, of those dying from cancers of the stomach and lung with those of all matching survivors. The RMLT analyses provided evidence of an occupational involvement in the risk of death from lung cancer from work in the foundry area or fettling shop, and weaker evidence of an occupational involvement in the risk of death from stomach cancer from work in the foundry area.