As part of the effort to establish industrial practice and public policy regarding asbestos in Zimbabwe, we have conducted a cross-sectional study of the chrysotile mines and mills. A stratified random sample of workers with greater than 10 years of exposure has been evaluated by spirometry, chest radiographs, and employment history. The latter was converted to quantitative estimates of exposure dose, using a matrix based on measured and reconstructed fiber levels for each job and facility during the years of work. Based on these data, a clear dose-response between asbestos exposure and functional loss has been demonstrated, with mean losses from predicted of about 400-600 cc in vital capacity in the 10% of the population with heaviest exposures. Low-grade parenchymal radiographic abnormalities (ILO grade greater than or equal to 1/0) were evident in 8.7% of the total study group and were almost 10 times more common in those with more than 100 fibers/cc.years cumulative exposure than in those with 16 fibers/cc.years or less. Pleural disease was relatively rare, occurring in just under 10% of the study group, and was unrelated to exposure dose. Overall, these findings are compatible with results of similar studies in Quebec and Swaziland and suggest that similar control strategies are probably indicated.