Plantwide analyses of the mortality experience of 8147 foundrymen revealed excesses for several diseases including lung cancer. Using indirect measures of smoking, it appeared that most, if not all, of the excess of lung cancer deaths could be explained by smoking habits. To explore further the possible association between these mortality excesses and foundry exposures, jobs were grouped into six work areas on the basis of similarities in production processes. The findings of analyses by work areas support the inferences from plantwide observations. No evidence was found of a relationship between lung cancer and foundry exposures. The pattern of mortality from emphysema and cerebrovascular disease in the different work areas paralleled that of lung cancer, suggesting that mortality from these diseases may have been influenced by a common etiologic agent, probably tobacco smoke. The data also reveal possible associations between metal pattern-making and colon cancer, silica or metal dust and stomach cancer, and carbon monoxide and ischemic heart disease.