Agreement between experimental data on maximum morphological damage and maximal predicated tissue does offers promise that mathematical dosimetry models can contribute to a better understanding of effective doses in animal studies and their correspondence to human exposure levels. Modeling must be a dynamic process to take advantage of new information and evolving conceptualizations of physical, chemical, and biological processes. Qualitative extrapolations from animals to man have been indirectly in setting National Ambient Air Quality Standards. Continued development of dosimetry models, along with information on species differences in sensitivity to pollutants, have the potential to make extrapolations between animals and man more quantitative in the future.