To determine the effect of long-term salmeterol aerosol therapy on airway hyperresponsiveness measured by methacholine challenge. Randomized, double-blind, placebo-controlled, multicenter study. Thirty-one clinical centers in the United States. Four hundred eight asthmatic patients > or = 12 years of age with baseline FEV1 of > or = 70% of predicted values. Patients were not using inhaled corticosteroids. Twice-daily salmeterol aerosol, 42 microg, or placebo via metered-dose inhaler for 24 weeks. Backup albuterol was available. Pulmonary function tests were performed before, during, and after treatment. Subjects recorded asthma-related symptoms, morning and evening peak expiratory flow (PEF) levels, and use of supplemental albuterol daily on diary cards. Methacholine challenges were performed 10 to 14 h postdose at weeks 4, 12, and 24, and 3 and 7 days posttreatment. Over 24 weeks of treatment, salmeterol provided significant (p < 0.001) protection against methacholine-induced bronchoconstriction of approximately one doubling dose of methacholine when compared to placebo with no evidence for a progressive decrease in protection. A rebound increase in airway hyperresponsiveness was not observed 3 and 7 days after cessation of salmeterol therapy. Salmeterol treatment resulted in sustained improvements of 0.21 to 0.26 L in morning premedication FEV1 and an improvement of 26.2 L/min in morning PEF when compared to placebo (p < 0.001). The use of salmeterol significantly reduced combined daytime asthma symptoms by 20% when compared to placebo (p = 0.005). A total of 34 and 48 exacerbations, respectively, were reported in the Salmeterol and placebo groups, and no evidence was present for a difference in the severity of asthma exacerbations between groups. Adverse event profiles were similar for the salmeterol and placebo groups. Regular long-term use of salmeterol aerosol resulted in sustained improvements in pulmonary function and asthma symptom control over the 24-week treatment period. There was no increase in bronchial hyperresponsiveness or loss of bronchoprotection at 24 weeks from that seen following 4 weeks of therapy. There was no evidence of rebound airway hyperresponsiveness after cessation of salmeterol treatment. Regular treatment with the long-acting beta-agonist salmeterol does not lead to clinical instability or vulnerability to unpredictable asthma attacks.