A mortality study of 3,328 South Dakota gold miners exposed to silica (14808607) and nonasbestiform amphibole minerals was updated by an extension of the follow up through 1990. The miners had worked underground for at least 1 year between 1940 and 1965. To estimate exposures to dust in the mine over time, a job/exposure matrix was created; full time underground jobs were divided into five major groups based on job function and dust exposures: laborers, miners, motormen, supervisors, and skip loaders. A sixth category was created for nonexposed workers. Existing measurements for each year from 1937 to 1975 were used for the calculation of average dust exposures for each job category. Analyses were conducted for mortality, life table, person years, national rates for multiple causes of death, lung cancer, and cigarette smoking factors. There were 1,551 deaths observed; silicosis was mentioned on death certificates for 140. The calculated standardized mortality ratio (SMR) for lung cancer was only marginally elevated for the cohort (SMR was 1.13 with US population as a referent group); a positive exposure response trend was not seen with lung cancer mortality and estimated cumulated dust exposure. In a 1960 cross sectional survey, miners were found to smoke more than the general US population. Other diseases associated with silica exposure were significantly increased; the SMRs for tuberculosis and silicosis were 3.44 and 2.61, respectively, with clear exposure response trends. Significant excesses of arthritis, musculoskeletal diseases, autoimmune conditions, and diseases of the blood and blood forming organs associated with silica exposure were observed in multiple cause analysis. The authors conclude that excess mortality from several causes of death known to be related to silica was observed in this cohort. Although lung cancer showed some elevation, a positive exposure/response trend was not seen with cumulative dust exposure. The discrepancy in the findings of this study compared to others may arise from differences in the mineralogic quality of the silica and the presence of confounding factors.