Longitudinal lung function data from four occupational health surveys was used to explore the relationship between nonspecific bronchial responsiveness (NSBR) and the rate of decline of FEV1 (RDFEV) and to address other factors that may predict or influence RDFEV. Of the 1,203 subjects with baseline methacholine and lung function data, follow-up data were available for 733 individuals (61%). The data-available and data-unavailable groups were well matched with respect to baseline lung function, atopy, and smoking status. Compared with the unavailable group, those available for follow-up were younger (42.5 versus 37.7 yr; p less than 0.0001) and slightly less responsive to methacholine (p less than 0.01). Somewhat unexpectedly, RDFEV was lower in the 30 asthmatic subjects than in the nonasthmatics. Among the nonasthmatic subgroup (96% of those available for follow-up), age, occupational group, and baseline FEV1 (% predicted) were independent predictive factors for RDFEV. When these factors were included in the model, RDFEV was found to be increased among current smokers compared with never-smokers or ex-smokers. In the final regression model, a relationship was found between methacholine sensitivity (calculated as a dose-response slope) and RDFEV in nonasthmatics (p less than 0.05). Stratification by smoking status revealed that the relationship between RDFEV and bronchial responsiveness was confined to current smokers and that atopy was an additional risk factor in this subgroup only. This relationship was valid among current smokers in each of the three occupational groups studies. However, reinclusion of the 30 asthmatic subjects in the study population obscured the relationship between NSBR and RDFEV.(ABSTRACT TRUNCATED AT 250 WORDS)