The review puts into perspective the relationship of human morbidity to the occurrence of sulfur oxides and related particulates in the atmosphere. There is no question that such a relationship exists. But the many problems of terminology and technology make a precise and unequivocal interpretation of the data impossible. Most of the morbidity data come from epidemiologic studies; air pollution epidemiology is undeniably a multidisciplinary endeavor with all the problems inherent therein; its interpretation, when recognizing these problems, can provide a holistic view that is greater than the sum of its components. Any single study in air pollution epidemiology could lead equally qualified, honest individuals to widely divergent interpretations. The data must, therefore, be interpreted in a critical, but open-minded fashion that recognizes methodologic problems and leans toward qualitative conclusions rather than straining for rigid quantitative results. This evidence provides a basis into which can be inserted the more quantitative data from animal toxicology, in vitro metabolic studies, and clinical research that lead to the final judgmental process of developing regulatory policy.