We assessed the relative impact of residential exposure to community air pollution and habitual cigarette smoking on lung function by comparing the annualized rate of change in forced expiratory volume in 1 s (FEV1) in current, former, and never-smokers 25 to 59 yr of age residing in three demographically similar areas of the Southern California air basin who had been chronically exposed to (1) moderate levels of photochemical oxidants and very low levels of other pollutants (Lancaster); (2) very high levels of photochemical oxidants, sulfates, and particulate matter (Glendora); and (3) high levels of sulfates, oxides of nitrogen, and probably hydrocarbons (Long Beach), together with moderate levels of sulfur dioxide. Of the 621 to 763 nonsmokers, 317 to 479 former smokers and 472 to 691 continuing smokers residing in the three areas who were studied initially, 53 to 64, 49 to 59, and 43 to 54%, respectively, were retested. For male residents, area of residence and smoking category each had highly significant effects on FEV1 decline (two-way ANCOVA; p < 0.001) without significant interaction (p > 0.4). After adjustment for baseline FEV1, age, height, and a history of allergy, the mean decline in FEV1 attributable to living in Long Beach compared with living in Lancaster was 23.6 ml/yr, which was 71% of the rate of decline in FEV1 (33.3 ml/yr), attributable to smoking > 1 pack of cigarettes per day. For female residents, a significant interaction was noted between area and smoking (p < 0.05).(ABSTRACT TRUNCATED AT 250 WORDS)