Holoprosencephaly (HPE), the most common developmental defect of the forebrain and midface, is caused by a failure to delineate the midline in these structures. Both genetic and environmental etiologies exist for HPE, and clinical presentation is highly variable. HPE occurs in sporadic and inherited forms, and even HPE in pedigrees is characterized by incomplete penetrance and variable expressivity. Heterozygous mutations in eight different genes have been identified in human HPE, and disruption of Sonic hedgehog expression and/or signaling in the rostroventral region of the embryo is a major common effect of these mutations. An understanding of the mechanisms whereby genetic defects and teratogenic exposures become manifest as developmental anomalies of varying severity requires experimental models that accurately reproduce the spectrum of defects seen in human HPE. The mouse has emerged as such a model, because of its ease of genetic manipulation and similarity to humans in development of the forebrain and face. HPE is generally observed in mice homozygous for mutations in orthologs of human HPE genes though, unlike humans, rarely in mice with heterozygous mutations. Moreover, reverse genetics in the mouse has provided a wealth of new candidate human HPE genes. Construction of hypomorphic alleles, interbreeding to produce double mutants, and analysis of these mutations on different genetic backgrounds has generated multiple models of HPE and begun to provide insight into the conundrum of the HPE spectrum. Here, we review forebrain development with an emphasis on the pathways known to be defective in HPE and describe the strengths and weaknesses of various murine models of HPE.