The association of age, gender, number of eosinophils, area of residence, cigarette smoking, respiratory symptom prevalence, and FEV1 with the level of bronchial responsiveness was studied in a random sample of 2,216 subjects aged 15 to 72 yr. Subjects participated in the Dutch longitudinal study on chronic obstructive pulmonary disease. In 18 yr of follow-up, 5,012 observations were collected. Interviewers used a standardized questionnaire to assess the presence of respiratory symptoms. Bronchial responsiveness was measured by a histamine challenge test. Because multiple measurements within a subject are correlated, multivariate regression methods for correlated outcome were used. A greater number of eosinophils, skin test positivity, and living in a rural area (Vlagtwedde) were associated with increased responsiveness, independently of the level of FEV1 and the presence of respiratory symptoms. Older age was associated with increased responsiveness, and this was even more so in subjects with symptoms. Cigarette smokers were more responsive than nonsmokers, but this association was not significant if the level of FEV1 was taken into account. Hyperresponsiveness was more likely to be present if the amount of cigarettes smoked per day was greater. The level of responsiveness did not differ significantly between males and females. For the same degree of obstruction, however, expressed as the FEV1/VC ratio, males tended to be less responsive than females. The analyses were repeated using the dose-response slope as a continuous measure of responsiveness and by applying a method to adjust for censoring the responsiveness data. These analyses yielded identical results.