Methods for reconstructing historic dust exposures, taking into account jobs and controls, were developed and combined with a retrospective cohort mortality study of chrysotile asbestos textile workers, between 1940 and 1975, to evaluate dose-response relationships for lung cancer and nonmalignant respiratory diseases. Among 768 white males, statistically significant excess mortality was observed for lung cancer (SMR=348) and nonmalignant respiratory diseases (SMR=263). There were 15 deaths due to asbestosis or pulmonary fibrosis. Using estimated fibre dose by job and calendar time, combined with detailed work histories, mortality by cumulative dose was investigated. For lung cancer, a linear dose-response relationship was obtained with an SMR of 223 in the lowest cumulative dose category. Significant dose-response relationships were also observed for nonmalignant respiratory diseases excluding acute upper respiratory infection, influenza, pneumonia and bronchitis. Of the 191 deaths among white males, only one was due to mesothelioma.