The collision of Indian and Eurasian landmasses in the Cenozoic was a decisive factor in shaping biodiversity patterns in Southern and Southeastern Asia. While most studies thus far have focused on the biotic interchange between India and Eurasia and evolutionary diversification on or around the Tibetan Plateau little attention has been paid to the biodiversity build-up in the Eastern Himalaya biodiversity hotspot (EHH) which harbours over 540 freshwater fish species with a high degree of endemicity. An important component of the regional ichthyofauna are snakeheads fishes of the family Channidae comprising throughout their African-Asian distribution 47 valid species, but a poorly known species-level diversity. In order to evaluate channid intrarelationships and biogeography a temporal and geographical framework of channid evolution in conjunction with a critical reevaluation of the channid fossil record is warranted. Based on molecular data, we provide a comprehensive species-level phylogeny based on 223 channid individuals belonging to 37 species and one additional currently undescribed species. The first split within channids separates the African genus Parachanna from the Asian genus Channa which can be divided into eight distinct species groups (C. micropeltes, C. lucius , C. asiatica, C. argus, C. marulius, C. striata, C. punctata and C. gachua species groups). Large intraspecific divergences were observed within several species and potentially indicate additional species-level diversity. Almost 40% of the channid species are narrow-range endemics belonging to the C. gachua species group. These are found in the EHH making this area an outstanding hotspot for endemic channid diversity. The large majority of the EHH endemics are restricted to the southern foothills of the Eastern Himalaya and the Shillong-Mikir Hills Plateau, areas west of the Indo-Burman Ranges. Our results reveal complex and difficult to interpret biogeographic patterns indicating that both vicariance and dispersal events have potentially been responsible in shaping current distribution patterns in Asian channids. We recognize †Parachanna fayumensis as the oldest reliable channid fossil, and argue that the three oldest so-called channid fossils (i.e., †Eochanna chlorakkiensis, †Anchichanna kuldanensis, and †Ophiocephalus lydekkeri) lack clear diagnostic features that would allow them to be unequivocally placed within Channidae.