The dose-response relationship for asbestos exposure in a chrysotile product factory was studied. The past gravimetric dust concentration values, obtained from different worksites, were converted into fiber concentration values according to conversion factors that were worked out by simultaneous sampling in this study. The conversions were made so that exposure could be expressed in fiber-years (f-yr). Asbestosis was diagnosed on the basis of chest radiographs and occupational histories. Cumulative dust exposure (f-yr) was calculated up to the date of diagnosis for asbestosis patients, and up to September 1982 for the remaining workers. A dose-response relationship expressed as fiber-years exposed vs cumulative prevalence of asbestosis was established by the life table method on the basis of these data. Predicted 3 and 1% prevalence of asbestosis corresponded to 43 and 22 f-yr exposure, respectively. Considering that a worker can work for 35 years, these doses are commensurate with dust concentrations of 1.22 and 0.63 f/ml, respectively. It is recommended that 1 f/ml be taken as the maximum allowable concentration of airborne asbestos dust for the workplace with an anticipated prevalence of about 2% asbestosis after 35 years of exposure.