A multicompartmental mathematical model has been used to simulate variations in the cytotoxicity of dusts in the kinetics of the retention, in the pulmonary region and tracheobronchial lymph nodes, of practically insoluble quartzite and titanium dioxide dust particles deposited on the free surfaces of the acini from alveolar air. Experiments with these dusts were conducted on rats exposed to virtually the same dust concentrations in the air for an experimental period of 20 weeks and a period of 10 weeks after exposure. Satisfactory approximation to the experimental data on the retention of these dusts is obtained by using the model parameters that depend either on damage to lung macrophages by phagocytosed particles or on the response of the host organism to this damage by enhanced recruitment of neutrophilic leucocytes; all the other variables of the model being unchanged. The values of the "action integral" computed from this model and multiplied by the index of comparative cytotoxicity of particles in vitro satisfactorily approximate to quantitative differences in the intensity of pneumoconioses caused by the dusts under study by the end of the experimental period. On the whole, the results of the mathematical model agree with the hypothesis that the cytotoxicity of particles plays a key part in both the process of retention of dust in the lung parenchyma and lung associated lymph nodes, and the pathological process caused by the retained dust. Thus given the factors and conditions on which the deposition of practically insoluble dusts in the pulmonary region depends, it is necessary to take into account the multiplicative nature of these two effects of cytotoxicity when predicting the comparative risk of pneumoconiosis. (Occup Environ Med 1994;51:173-180) The breakdown of lung macrophages as a result of ingesting poorly soluble dust particles is a key link in the pathogenesis of pneumoconioses. This link determines both the effectiveness of lung clearance (and thus the kinetics of retention of dust in the lung parenchyma and lung associated lymph nodes) and the intensity of pathological changes in these organs caused by long term dust retention. Most of the researchers in this field share this view, with subtle differences. It still remains, however, a hypothesis that has not been treated quantitatively. The dual effect of cytotoxicity of dust is not disputed, but the statement that it has a key role still remains a speculative one.