Summary Background Some studies suggest that patients with asthma who are homozygous for arginine at the 16th aminoacid position of the β 2-adrenergic receptor (B16 Arg/Arg) benefit less from treatment with longacting β 2 agonists and inhaled corticosteroids than do those homozygous for glycine (B16 Gly/Gly). We investigated whether there is a genotype-specific response to treatment with a longacting β 2 agonist in combination with inhaled corticosteroid. Methods In this multicentre, randomised, double-blind, placebo-controlled trial, adult patients with moderate asthma were enrolled in pairs matched for forced expiratory volume in 1 s and ethnic origin, according to whether they had the B16 Arg/Arg (n=42) or B16 Gly/Gly (n=45) genotype. Individuals in a matched pair were randomly assigned by computer-generated randomisation sequence to receive inhaled longacting β 2 agonist (salmeterol 50 μg twice a day) or placebo given in a double-blind, crossover design for two 18-week periods. Open-label inhaled corticosteroid (hydrofluoroalkane beclometasone 240 μg twice a day) was given to all participants during the treatment periods. The primary endpoint was morning peak expiratory flow (PEF). Analysis was by intention to treat. This trial is registered with ClinicalTrials.gov, number NCT00200967. Findings After 18 weeks of treatment, mean morning PEF in Arg/Arg participants was 21·4 L/min (95% CI 11·8–31·1) higher when participants were assigned to receive salmeterol than when assigned to receive placebo (p<0·0001). In Gly/Gly participants, morning PEF was 21·5 L/min (11·0–32·1) higher when participants were assigned to receive salmeterol than when assigned to receive placebo (p<0·0001). The improvement in PEF did not differ between genotypes (difference [Arg/Arg–Gly/Gly] −0·1, −14·4 to 14·2; p=0·99). In Gly/Gly participants, methacholine PC 20 (20% reduction in forced expiratory volume in 1 s; a prespecified secondary outcome) was 2·4 times higher when participants were assigned to salmeterol than when assigned to placebo (p<0·0001). Responsiveness to methacholine did not differ between salmeterol and placebo in Arg/Arg participants (p=0·87). The 2·5 times higher genotype-specific difference in responsiveness to methacholine was significant (1·32 doubling dose difference between genotypes, 0·43–2·21, p=0·0038). Seven Arg/Arg participants (placebo, n=5; salmeterol, n=2) and six Gly/Gly participants (placebo, n=3; salmeterol, n=3) had an asthma exacerbation. Five serious adverse events were reported, one each during the pre-match and run-in phases on open-label inhaled corticosteroid, two during double-blind treatment with salmeterol/inhaled corticosteroid, and one during double-blind treatment with placebo/inhaled corticosteroid. None of the serious events was asthma-related or related to study drugs or procedures. Interpretation In asthma patients with B16 Arg/Arg and B16 Gly/Gly genotypes, combination treatment with salmeterol and inhaled corticosteroid improved airway function when compared with inhaled corticosteroid therapy alone. These findings suggest that patients should continue to be treated with longacting β 2 agonists plus moderate-dose inhaled corticosteroids irrespective of B16 genotype. Further investigation is needed to establish the importance of the genotype-specific difference in responsiveness to methacholine. Funding National Institutes of Health.