Intravenous lidocaine can attenuate bronchial hyperreactivity. However, lidocaine inhalation might yield the same or better results at higher airway and lower lidocaine plasma concentrations. Therefore, we tested in awake volunteers with bronchial hyperreactivity the effect of lidocaine on histamine-induced bronchoconstriction administered either intravenously or as an aerosol. After approval of the local ethics committee, 15 volunteers were enrolled in this placebo-controlled, double-blinded, randomized study. Volunteers were selected by showing a decrease in FEV1 greater than 20% of baseline (PC20) in response to histamine inhalation. On three different days the challenge was repeated after pretreatment with either intravenous lidocaine, inhaled lidocaine, or placebo. Blood samples for determination of lidocaine plasma concentration were drawn. Comparisons were made using the Friedman and Wilcoxon signed-rank tests. Baseline PC20 was 6.4 +/- 1.1 mg. ml-1. Both inhalation of lidocaine and intravenous administration significantly increased PC20 to 14.8 +/- 3.5 mg. ml-1 and 14.2 +/- 2. 5 mg. ml-1, respectively (p = 0.0007). Peak plasma lidocaine concentrations at the end of challenges were 0.7 +/- 0.1 microg. ml-1 (inhaled) and 2.2 +/- 0.1 microg. ml-1 (i.v.). However, 7 of 15 subjects showed an initial decrease of FEV1 greater than 5% following lidocaine inhalation. While both intravenous as well as inhaled lidocaine attenuate reflex bronchoconstriction significantly, lidocaine plasma concentrations are significantly lower after inhalation. However, the high incidence of initial bronchoconstriction to lidocaine inhalation may limit its use in patients with asthma and thus offers therapeutic advantages for intravenous lidocaine.