Measurements of bronchial caliber and airway sensitivity were performed 4 times during the day (at 9, 11, 16, and 22 hr) at basal conditions (baseline), following the first inhalation of 50 micrograms salmeterol (acute) and at the 21st, 90th and 150th day after the initiation of an uninterrupted long-term treatment with inhaled salmeterol (50 micrograms b.i.d., at 10 and 22 hr). In each period of the protective effect was assessed by computing the increase of the methacholine dose able to induce a 20% fall of the forced expiratory volume in the first second (PD20FEV1) in terms of doubling dose (DD), either against the respective 9-hour PD20FEV1 value (DD9hr) or against the corresponding baseline PD20FEV1 value (DDbaseline). After the first dose of salmeterol the forced expiratory volume in the first second (FEV1) increased significantly as compared with the 9-hour FEV1 and the corresponding baseline FEV1 at each observation time (p < 0.01). During regular treatment FEV1 was higher than baseline at the 21st and 90th day at each observation time (p < 0.05), whereas at the 150th day no significant FEV1 increments were observed at 9 hr and 22 hr. The acute protective effect exerted by salmeterol amounted to about 2 DD9hr (p < 0.05) and 2 DDbaseline (p < 0.05) at each observation time. At the 21st, 90th, and 150th day, however, no significant increase of DD9hr was found, although a mild decrease of airway sensitivity of 1 DDbaseline of magnitude was observed for all periods at each observation time. We conclude that in mild to moderate asthma salmeterol appears to rapidly lose its ability to improve bronchial responsiveness while it is effective in maintaining a well-sustained bronchodilation despite a small degree of tachyphylaxis.