The long-acting beta 2-agonist salmeterol has been shown in several in vitro studies to produce non-beta-mediated relaxant effects. The aim of the present study was to investigate whether these effects have any relevance in humans in vivo. Thirteen healthy individuals were studied in a randomized, double-blind, cross-over study on five separate days. The subjects were pre-treated orally with either propranolol 400 mg in order to block beta-adrenoceptor mediated effects or placebo. Two hours after drug intake, three increasing doses of salmeterol (25 + 50 + 100 micrograms), salbutamol (100 + 200 + 400 micrograms) or placebo were given from matched meter dose inhalers at 1-h intervals between doses. Specific airway conductance (sGAW) was measured in a body plethysmograph at the beginning of the experiment and 30 and 60 min after each inhaled dose of the beta-agonists. Salmeterol and salbutamol produced the same maximal increase in sGAW and had the same area under the dose-response curves. Pre-treatment with propranolol totally inhibited the effect of both drugs. In conclusion, salmeterol at clinically used doses did not produce any non-beta-mediated bronchodilating effect in normal individuals, measured as sGAW. Salmeterol and salbutamol showed the same efficacy but salmeterol was four times more potent than salbutamol.