OBJECTIVES To study whether the mortality from non-malignant and malignant respiratory diseases of workers employed in French and Austrian talc mines and mills is related to their long term occupational exposure. METHODS Two historical cohorts were set up comprising all male subjects who had been working continuously for at least 1 year in a series of talc producing companies in France and Austria. The French cohort consisted of those employed at a site in the French Pyrenees and working between 1 January 1945 and 31 December 1994. The Austrian cohort consisted of the workers employed between 1 January 1972 and 31 December 1995 in one of four industrial sites in the Austrian Alps. The mortality within the cohorts was compared with local death rates. Two nested case-control studies focusing on non-malignant and malignant respiratory diseases were set up to estimate possible dose-response relations with cumulative exposure to talc dust based on an industry specific job exposure matrix. RESULTS Mortality from lung cancer was in small excess in both cohorts (France, standardised mortality ratio (SMR) 1.23, 21 cases observed, 95% confidence interval (95% CI) 0.76 to 1.89; Austria, SMR 1.06, seven observed, 95% CI 0.43 to 2.19). A non-significant excess mortality was found for all non-malignant respiratory diseases in the French cohort due to a significant excess for pneumoconiosis (SMR 5.56, three observed, 95% CI 1.12 to 16.2). The case-control study of non-malignant respiratory disease showed an increased mortality in the highest exposure groups (odds ratio (OR) 2.5 for a cumulative exposure > or = 800 y.mg/m(3)) with a significant trend (OR/100 y.mg/m(3) 1.08) with cumulative exposure to talc. On the contrary, no increasing trend could be found in the case-control study of lung cancer. This result must be interpreted considering the small cohort size. Adjustment on smoking and exposure to quartz did not influence these results to any extent. CONCLUSIONS The mortality from non-malignant respiratory disease was found to be related to high cumulative exposure to talc dust. The small excess in lung cancer does not seem to be attributable to talc.