Studies on the fibrogenicity of asbestos (1332214) and other mineral particles retained in human lungs were summarized. Lung tissue specimens from South African miners who had been employed in mines that produced a small diameter crocidolite (12001284) or mines that produced crocidolite and amosite (12172735) were examined to determine the extent of fibrosis and fiber retention. A mathematical model was developed to determine which fiber parameter, such as number concentration, surface area, or volume, would show equivalent asbestos doses, as measured by degree of fibrosis. Tissue samples showing the same degree of fibrosis were considered to have received the same asbestos dose. The relevant parameter was determined to be total surface area of the retained fiber per unit weight of tissue. The degree of fibrogenicity was independent of asbestos type. Lung tissue specimens obtained at autopsy from miners who had been employed at four asbestos mines in South Africa were analyzed for their asbestos content and the degree of fibrogenicity was assessed. The tissue specimens contained chrysotile (12001295), quartz (14808607), crocidolite, amosite, and anthophyllite (17068789). Chrysotile and quartz appeared to show fibrogenicities that were similar to that of crocidolite, amosite, and anthophyllite. Lung tissue samples were taken at autopsy from persons who had been employed at asbestos mines and factories, gold mines, a platinum mine, shipyards, and other dusty workplaces. The mineral content, fiber dimensions, and degree of fibrosis of the specimens were determined. The data were treated by logistic regression techniques to determine which fiber parameters were best correlated with fibrogenicity. Aggregated surface area of all particles, not just asbestos fibers, was the best predictor of fibrogenicity. Aggregated particle volume also provided a reasonable index of fibrogenic behavior. Particle number concentration was a poor predictor of fibrogenicity.