Background: Long-acting β2-agomsts (LABA) and leukotnene receptor antagonists (LTRA) are two principal agents that can be added to inhaled corticosteroids (ICS) for patients with asthma that is not adequately controlled by ICS alone. In our previous study, the Gly16Arg genotype of the β2-adrenergic receptor (ADRB2) gene did not influence the differential bronchodilator effect of salmeterol versus montelukast as an add-on therapy to ICS within 16 weeks of follow-up (the J-Blossom study). Methods: We examined if genes encoding CYSLTR1, CYSLTR2, PTGER2 or PTGER4 could explain differential responses to salmeterol versus montelukast usmg the participants of the J-Blossom study. This study included 76 patients with mild-to-moderate asthma. The difference in peak expiratory flow (PEF) (ΔPEF, l/mm) after 16 weeks of treatment with salmeterol (ΔPEFsal) versus montelukast (ΔPEFmon) was associated with the genotypes at each of 4 genes. In addition, multivariate analyses were used to identify a gene-gene interaction between ADRB2 gene and each of these 4 genes. Results: Although none of 4 genes were associated with ΔPEFsal - ΔPEFmon in the univariate analyses, multivariate analysis showed that PTGER4 gene, interacting with ADRB2 Glyl6Arg, was associated with ΔPEFsal - ΔPEFmon (p=0.0032). Conclusion: Our findings suggested that the interactions between two genetic loci at ADRB2 and PTGER4 is important in determining the differential response to salmeterol versus montelukast in patients with chrome adult asthma.