Severe respiratory disease arising from influenza virus infection has a high fatality rate. Neutrophil myeloperoxidase (MPO) has been implicated in the pathogenesis of severe influenza-induced pneumonia because extracellularly released MPO mediates the production of hypochlorous acid, a potent tissue injury factor. To search for candidate anti-influenza compounds, we screened leucomycin A3 (LM-A3), spiramycin (SPM), an erythromycin derivative (EM900, in which anti-bacterial activity has been eliminated), and clarithromycin (CAM), by analyzing their ability to inhibit MPO release in neutrophils from mice and humans. When each candidate was injected into mice infected with a lethal dose of A/H1N1 influenza virus (PR-8), LM-A3 produced the highest survival rate (80.9%). We found that LM-A3 induced beneficial effects on lung pathology and viral proliferation involved in the regulatory activity of MPO release, pro-inflammatory cytokines and interferon-α production in the lung. SPM and EM900 also induced positive survival effects in the infected mice, whereas CAM did not. We further found that these compounds inhibit virus proliferation in human pneumonia epithelial A549 cells in vitro. LM-A3 showed effective action against influenza A virus infection with high anti-viral activity in human host cells, indicating the possibility that LM-A3 is a prospective lead compound for the development of a drug for human influenza. The positive survival effect induced by EM900 suggests that pharmacological architectures between anti-bacterial and anti-influenza virus activities can be dissociated in macrolide derivatives. These observations provide valuable evidence for the potential development of novel macrolide derivatives that have strong anti-viral but no anti-bacterial activity.