In a random population sample of 1905 subjects we studied the occurrence of respiratory symptoms in relation to airways responsiveness. Responders (PC10 FEV1 to histamine at 16 mg.ml-1 or less) had crude prevalence rates two to three times higher than nonresponders. In logistic regression analysis, odds ratios were estimated for each threshold value, compared to the reference value (greater than 32 mg.ml-1), controlling for age, sex, area of residence, and smoking habit. Odds ratios increased with decreasing threshold values in a dose-response relationship for all symptoms, except for bronchitis periods. We analysed the association of airways responsiveness to pulmonary function level by multiple linear regression, controlling for age, sex, height, area of residence, and smoking habit. There was an inverse relationship of FEV1 level to threshold value. Male subjects within a threshold value of 1 mg.ml-1 had a mean adjusted FEV1 of 1170 ml less than males with a threshold value of greater than 32 mg.ml-1. The relationship of responsiveness to decline of FEV1 with time was studied in 186 male subjects who took part in five consecutive surveys from 1967 to 1981. The greatest mean adjusted yearly decline was noted in responding smokers: 35.3 ml per yr, compared to nonreactive nonsmokers: 10.9 ml per yr. Regression analysis of the yearly decline in 169 subjects with at least two pairs of two consecutive threshold tests revealed that the more positive tests subjects had, the greater was their mean adjusted yearly decline. It is concluded that airways responsiveness may be an important factor in the development of chronic obstructive pulmonary disease.