Modulating the host response is a promising approach to treating influenza, a virus whose pathogenesis is determined in part by the reaction it elicits within the host. Though the pathogenicity of emerging H7N9 influenza virus has been reported in several animal models, these studies have not included a detailed characterization of the host response following infection. To this end, we characterized the transcriptomic response of BALB/c mice infected with H7N9 (A/Anhui/01/2013) virus and compared it to the responses induced by H5N1 (A/Vietnam/1203/2004), H7N7 (A/Netherlands/219/2003) or pandemic 2009 H1N1 (A/Mexico/4482/2009) influenza viruses. We found that responses to the H7 subtype viruses were intermediate to those elicited by H5N1 and pdm09H1N1 early in infection, but that they evolved to resemble the H5N1 response as infection progressed. H5N1, H7N7 and H7N9 viruses were pathogenic in mice, and this pathogenicity correlated with increased transcription of cytokine response genes and decreased transcription of lipid metabolism and coagulation signaling genes. This three-pronged transcriptomic signature was observed in mice infected with pathogenic H1N1 strains such as the 1918 virus, indicating that it may be predictive of pathogenicity across multiple influenza strains. Finally, we used host transcriptomic profiling to computationally predict drugs that reverse the host response to H7N9 infection, and identified six FDA-approved drugs that could potentially be repurposed to treat H7N9 and other pathogenic influenza viruses.