The genetic differentiation within a species in the Himalaya-Hengduan Mountains (HHM) region was considered as the combined consequence of historical orogenesis and climatic oscillations. The visualization of dispersal corridors as the function of population genetic connectivity became crucial to elucidate the spatiotemporal dynamics of organisms. However, geodiversity and physical barriers created by paleo-geoclimatic events acted vigorously to impact notable alterations in the phylogeographic pattern and dispersal corridors. Therefore, to achieve detailed phylogeography, locate dispersal corridors and estimate gene flow, we integrated phylogeography with species distribution modelling and Least Cost method of Mirabilis himalaica (Edgew.) Heimerl in the HHM. We amplified four cpDNA (petL-psbE, rps16-trnK, rps16 intron, trnS-trnG), and a low copy nuclear gene (G3pdh) sequences from entire 29 populations. SAMOVA, genealogical relationships, and phylogenetic analysis revealed four structured phylogroups for M. himalaica with the onset of diversification in late Pliocene c. 3.64 Ma. Bayesian skyline plot (BSP) indicated demographic or no recent demographic growth, which is supported by mismatch and neutrality tests. Paleo-distribution modelling revealed the range dynamics of M. himalaica to be highly sensitive to geo-climatic change with limited long-distance dispersal ability and potential evolutionary adaptation. Furthermore, river drainage systems, valleys and mountain gorges were identified as the corridors for population genetic connectivity among the populations. It is concluded that recent intense mountain uplift and subsequent climatic alterations including monsoonal changes since Pliocene or early Pleistocene formulated fragmented habitats and diverse ecology that governed the habitat connectivity, evolutionary and demographic history of M. himalaica. The integrative genetic and geospatial method would bring new implications for the evolutionary process and conservation priority of HHM endemic species.