A study of 2,611 long-term asbestos insulators was well suited to provide information on (1) the prevalence of spirometric impairments in a large, well-defined population and (2) the effects of cigarette smoking, radiographic abnormalities, and duration from onset of exposure on pulmonary function. Prevalences are reported by a mutually exclusive classification of impairments (normal, restrictive, obstructive, small airways, and combined) as well as by abnormality of specific spirometric tests (FVC, FEV1/FVC, and midexpiratory time). Only 3 percent of nonsmokers (NS) had obstruction and 6 percent a decreased FEV1/FVC. Frequency of restriction did not vary by smoking history; it was 31 percent in NS and current smokers (CS) and 34 percent in ex-smokers (XS). Obstruction (present in 17 percent) and combined impairment (in 18 percent) were most common in CS. The FEV1/FVC was decreased in 35 percent of CS and 18 percent of XS. The FVC was decreased in 49 percent of CS, 44 percent of XS, and 33 percent of NS. Normal spirometry was most common when the radiograph was normal (almost half the workers with normal radiographs had normal spirometry). Nevertheless, FVC was reduced in 27 percent of those with normal radiographs and a normal radiograph was seen in 11 percent of workers with restriction. Restrictive and combined impairments were most frequent when both parenchyma and pleura were abnormal. Restriction was more frequent in isolated pleural disease (seen in 34 percent such subjects) than in isolated parenchymal disease (22 percent). The contribution of pleural fibrosis to reduced FVC and of asbestos exposure and smoking both to reduced FVC and to reduced FEV1/FVC was confirmed by regression analysis. That reduced FVC and reduced FEV1/FVC are both more frequent in insulators who have smoked (compared with NS insulators or smokers in the general population) suggests an interaction between asbestos and smoking in producing both these physiologic abnormalities.