The filoviruses marburg- and ebolaviruses can cause severe hemorrhagic fever (HF) in humans and nonhuman primates. Because many cases have occurred in geographical areas lacking a medical research infrastructure, most studies of the pathogenesis of filoviral HF, and all efforts to develop drugs and vaccines, have been carried out in biocontainment laboratories in non-endemic countries, using nonhuman primates (NHPs), guinea pigs and mice as animal models. NHPs appear to closely mirror filoviral HF in humans (based on limited clinical data), but only small numbers may be used in carefully regulated experiments; much research is therefore done in rodents. Because of their availability in large numbers and the existence of a wealth of reagents for biochemical and immunological testing, mice have become the preferred small animal model for filovirus research. Since the first experiments following the initial 1967 marburgvirus outbreak, wild-type or mouse-adapted viruses have been tested in immunocompetent or immunodeficient mice. In this paper, we review how these types of studies have been used to investigate the pathogenesis of filoviral disease, identify immune responses to infection and evaluate antiviral drugs and vaccines. We also discuss the strengths and weaknesses of murine models for filovirus research, and identify important questions for further study.