Several challenge procedures have been developed to characterize the cough reflex in patients with airway diseases. This study was performed to compare the interindividual range of cough sensitivity in asthmatic and normal subjects as well as smokers using an identical method. Sixteen normal subjects, 20 patients with mild bronchial asthma, 6 patients with moderate to severe bronchial asthma, 9 current smokers, and 7 occasional smokers were included. In all subjects, methacholine challenges and standardized citric acid challenges were performed. Sensitivity of the cough reflex was expressed as cough threshold, i.e., as concentration at which coughing occurred. Reproducibility was assessed in 23 subjects. Within a concentration range of 0.625-320.0 mg/ml, inhaled citric acid caused cough in all subjects. Geometric mean (range) cough threshold was 13 (2.5-160) in normal subjects, 14 (5-40) in patients with mild, and 32 (20-40) mg/ml in patients with moderate to severe asthma, 40 (20-80) in current smokers, and 119 (80-160) in occasional smokers. Cough thresholds were reproducible within one doubling concentration. In normal subjects and patients with mild bronchial asthma, thresholds were not significantly different from each other but lower than those of the other groups (p<0.05 each). Cough thresholds in smokers and patients with moderate to severe asthma did also not differ significantly and were lower than in occasional smokers (p<0.05). There was no significant correlation between cough threshold, baseline FEV subset1 , and methacholine responsiveness. Our data indicate that (1) subjects with mild asthma showed on average similar cough thresholds as normal subjects, (2) there was a large variation in cough thresholds within groups, (3) the reproducibility of cough thresholds was within one doubling concentration, (4) cough thresholds did not correlate with methacholine responsiveness or baseline airway tone. In view of the prevalence of cough as a symptom of bronchial asthma, it appears that the determination of citric acid-induced cough thresholds does not yield additional diagnostic information in these subjects.