Objective The objective of this study is to examine the exposure- response relationship between prolonged exposure to low levels of silica, uranium, radon progeny and criteria pollutants and the respiratory health of uranium workers. Method Initially two pilot studies were done on selected groups of high-risk workers to determine the prevalence of alpha-1-antitrypsin deficiency and the retained burden of uranium in the lungs of exposed workers. The value of sputum cytology was also evaluated. A cross-sectional study was conducted in 1984 to determine the prevalence of spirometric abnormalities among uranium miners. The data from non-smokers without symptoms was used to develop lung function reference values relevant for the present population. Follow up analytical cross-sectional and case control studies were done in 1996 and 1999 to determine the mean independent effect of uranium on the respiratory health of miners and to detect abnormalities in both smokers and non¬smokers. Data was collected on employment history, smoking habits, pulmonary function and chest radiographs. To determine the independent mean effect of uranium mining on the respiratory health of workers, only data of non-smokers in exposed cohorts were compared with unexposed control cohorts. Results The prevalence of alpha-1-antitrypsin appeared to be high in workers with abnormal spirometric indices. The lung burden of retained uranium in workers in the final product recovery area was within normal limits. The sputum cytology was abandoned because it was found worldwide that cytological screening had no impact on population survival rates. Lung function prediction equations derived from multiple linear regressions were calculated for the study population which differed from Schoenberg et al prediction equations. The data from the analytical cross-sectional study in1996 suggests that the level of exposure in this study group (without the confounding effect of cigarette smoking) is not associated with statistically significant impairment of lung function, increased prevalence of silicosis, tuberculosis and respiratory cancer. Only the FEF75 was found to be an indicator of early lung abnormality. Data from the 1999 study supports the hypothesis that there is a small but significant exposure-response relationship between prolonged exposure to low levels of silica dust, and lung function abnormalities (in the absence of radiologically diagnosed silicosis). The prevalence of chronic obstructive pulmonary disease (COPD) and small airways disease (SAD) was small but significantly higher in exposed workers. The risk of developing COPD was 2.7 times higher for exposed workers. The risk for non-smokers was small but significant. The incidence of tuberculosis and lung cancer was not higher than expected (probably because ex-employees were not fully investigated). Conclusion Calculation of location specific prediction equations, for different ethnic groups is advocated to enable the earliest detection of lung function abnormalities. FEF75 could be used as an early indicator of lung function impairment (provided strict quality control is maintained). Exposure to a combination of silica dust, radon progeny and smoking was associated with an increased risk of lung function abnormalities, but not with tuberculosis, silicosis and lung cancer. Conclusive assessment must await sufficient latency, and ex-employees must be investigated to establish the true magnitude and distribution of the effects of siliceous and uranium dust on the health of workers (which may continue after exposure cease).