Regular beta(2)-adrenoceptor agonist therapy may lead to a rebound increase in bronchial responsiveness on discontinuation of therapy and a reduction in bronchoprotective effects. Formoterol, a long-acting beta(2)-agonist, is effective in single doses in the prevention of methacholine-induced bronchoconstriction. In a double-blind, placebo-controlled cross-over study, we examined the effect of an inhaled long-acting beta(2)-adrenoceptor agonist, formoterol (24 mu g twice a day) for 2 wk on airway function and responsiveness in 17 subjects with mild asthma (mean age, 26.3 +/- 1.4 yr) who were not taking inhaled glucocorticosteroids. FEV(1) and the provocative concentration of methacholine causing a 20% fall in FEV(1) (PC20) were measured at 36, 60, and 108 h and at 2 wk after the last dose of regular treatment. In addition, PC20 was measured 12 h after the first and the last dose of formoterol and placebo. PC20 values at 36, 60, 108 h and at 2 wk after formoterol treatment cessation were not significantly different from those after placebo. Mean FEV(1) was 3.44 +/- 0.18 L after placebo compared with 3.79 +/- 0.20 L after formoterol (p < 0.001) 12 h after the first dose, and mean PC20 was 0.53 (GSEM 1.4) mg/ml after placebo compared with 2.0 (GSEM 1.4) mg/ml after formoterol (p < 0.001). After 2 wk of regular treatment, mean FEV(1) at 12 h after the final dose of formoterol fell to 3.51 +/- 0.23 L compared with 3.41 +/- 0.18 L after the final dose of placebo (p = 0.03). Mean PC20 increased to 0.77 (GSEM 1.3) mg/ml after the final dose of formoterol, but this was a reduction in the protective effect of the first dose of formoterol compared with placebo (95% confidence interval for treatment difference: 0.19 to 0.95 mg/ml; p = 0.04). These observations indicate that formoterol can be discontinued after 2 wk of treatment without rebound hyperresponsiveness but that its bronchodilator and bronchoprotective effects are reduced.