We studied the tussive effects of a chloride-deficient solution (1.26% sodium bicarbonate). Nine normal volunteers and 10 mild asthmatic subjects were studied. In two double-blind, placebo-controlled, cross-over studies, we assessed the profile of any inhibitory effects that inhaled frusemide had over these responses. Baseline cough challenge was followed by inhalation of either frusemide (40 mg), or 0.15 M NaCl control. Cough was then induced at 0.5, 2, 4 and 6 h after treatment. Forced expiratory volume in one second (FEV1) was measured before and after each challenge. Changes from the baseline cough response due to drug or control were compared nonparametrically at each time point. There was no difference in the sensitivity of normal and asthmatic subjects to the cough challenge (median cough response 15 and 14.5 on control day, 12 and 15 on frusemide day). Frusemide caused sustained inhibition of the cough response in normal subjects (p < 0.05 at 2 h, p < 0.01 at 4 h), but had only a small, nonsignificant effect in asthmatic subjects at 30 min. Falls in FEV1 of asthmatic subjects due to the chloride-deficient solution were not significant, and did not correlate with number of coughs. We conclude that mild asthmatic subjects are less sensitive than normal subjects to the influence of frusemide against low chloride challenge. This observation is not explained by bronchoconstrictor effects of the cough challenge in asthmatic subjects.