Studies performed on airway smooth muscle in vitro have indicated that salmeterol is a partial agonist on the beta2-receptor in comparison to formoterol. In the present study we evaluated whether these pharmacological differences between salmeterol and formoterol also are applicable to asthmatic patients. The protective effects by increasing cumulative doses of formoterol (12, 60, 120 micrograms) and salmeterol (50, 250, 500 micrograms) on methacholine-induced bronchoconstriction were evaluated in a double-blind, crossover, placebo-controlled design. Patients were regularly treated with salbutamol 200 micrograms twice daily during the study period, to avoid variability in beta2-adrenoceptor tolerance. S-potassium, heart rate corrected Q-T interval (Q-Tc), and tremor score were followed as measures of systemic effects. Formoterol dose-dependently protected against methacholine responsiveness (4.6 doubling doses after 120 micrograms). Salmeterol, however, showed a flatter dose-response curve, and a significantly weaker maximal protective effect (2.8 doubling doses after 250 micrograms). Formoterol caused a significantly higher tremor score and a larger drop in S-potassium than salmeterol at the highest doses. These data show that salmeterol is a partial agonist on the beta2-receptor in relation to formoterol in human airways in vivo. Further studies are required to document the clinical consequences of this finding, for example in severe asthmatic patients.