We discovered a new lineage of the globally important fungal pathogen Cryptococcus gattii on the basis of analysis of six isolates collected from three locations spanning the Central Miombo Woodlands of Zambia, Africa. All isolates were from environments (middens and tree holes) that are associated with a small mammal, the African hyrax. Phylogenetic and population genetic analyses confirmed that these isolates form a distinct, deeply divergent lineage, which we name VGV. VGV comprises two subclades (A and B) that are capable of causing mild lung infection with negligible neurotropism in mice. Comparing the VGV genome to previously identified lineages of C. gattii revealed a unique suite of genes together with gene loss and inversion events. However, standard URA5 restriction fragment length polymorphism (RFLP) analysis could not distinguish between VGV and VGIV isolates. We therefore developed a new URA5 RFLP method that can reliably identify the newly described lineage. Our work highlights how sampling understudied ecological regions alongside genomic and functional characterization can broaden our understanding of the evolution and ecology of major global pathogens.