Otomops martiensseni is sparsely distributed throughout sub-Saharan Africa and southwestern Arabia (Yemen). Otomops madagascariensis from the dry portions of Madagascar is widely recognised to be a distinct species. Based on mitochondrial DNA sequences of the cytochrome b gene (1,004 base pairs; n = 50) and the control region (D-loop, 290 base pairs; n = 52), two Oriental outgroup species (O. wroughtoni and O. cf. formosus) formed a monophyletic clade that was the sister group to the Afro-Malagasy taxa, composed of O. martiensseni and O. madagascariensis. Within the Afro-Malagasy clade, we discovered three well-supported but genetically similar clades (inter-clade genetic distances of 3.4–4.4%) from 1) north-eastern Africa and Arabia, 2) African mainland except northeast Africa, and 3) Madagascar. Taken together, haplotype networks, estimated divergence times, regional species richness and historical demographic data tentatively suggested dispersal from Asia to Africa and Madagascar. To understand ecological determinants of phylogeographic, biogeographic and genetic structure, we assessed the potential distribution of O. martiensseni throughout sub-Saharan Africa with ecological niche modelling (MaxEnt) based on known point localities (n = 60). The species is predicted to occur mainly in woodlands and forests and in areas of rough topography. Continuity of suitable habitats supported our inferred high levels of continental gene flow (relatively low genetic distances), and suggested that factors other than habitat suitability have resulted in the observed phylogeographic structure (e.g., seasonal mass migrations of insects that might be tracked by these bats). Based on a Bayesian relaxed clock approach and two fossil calibration dates, we estimated that African and Oriental clades diverged at 4.2 Mya, Malagasy and African clades at 1.5 Mya, and African clades 1 and 2 at 1.2 Mya. Integrating phylogenetic, phylogeographic, population genetic and ecological approaches holds promise for a better understanding of biodiversity patterns and evolutionary processes.