Viruses usually exhibit strict species-specificity as a result of co-evolution with the host. Thus, in mouse models, a great barrier exists for analysis of infections with human-tropic viruses. Mouse models are unlikely to faithfully reproduce the human immune response to viruses or viral compounds and it is difficult to evaluate human therapeutic efficacy with antiviral reagents in mouse models. Humans and mice essentially have different immune systems, which makes it difficult to extrapolate mouse results to humans. In addition, apart from immunological reasons, viruses causing human diseases do not always infect mice because of species tropism. One way to determine tropism would be a virus receptor that is expressed on affected cells. The development of gene-disrupted mice and Tg mice, which express human receptor genes, enables us to analyze several viral infections in mice. Mice are, indeed, susceptible to human viruses when artificially infected in receptor-supplemented mice. Although the mouse cells less efficiently permit viral replication than do human cells, the models for analysis of human viruses have been established in vivo as well as in vitro, and explain viral pathogenesis in the mouse systems. In most systems, however, nucleic acid sensors and type I interferon suppress viral propagation to block the appearance of infectious manifestation. We herein review recent insight into in vivo antiviral responses induced in mouse infection models for typical human viruses.