The leafy liverwort genus Herbertus exhibits considerably variable morphology and widely disjunct distributions in both hemispheres. Here, the biogeographic history of the genus and its phylogenetic relationships with the focus on the taxonomically difficult, northern hemispheric disjunct species, were investigated. We conducted a time-calibrated, molecular-based phylogenetic analysis using psbA, trnL-F, and ITS1-2 loci and different approaches for ancestral range inference of the genus. Herbertus is inferred to have originated in the Cenozoic era about 51 million years ago, in an ancestral area including southern South America, the Neotropics, Oceania, and South-east Asia. The current distribution of the genus is supported to have resulted from long-term in situ persistence, short and long distance dispersals, extinctions and recolonizations. Postglacial range shifts between the southern and northern hemisphere in the genus show distinct patterns. In the southern hemisphere, Herbertus is characterized by in situ persistence, not showing further dispersal until the uplift of the Andean Cordillera. Species of the northern hemisphere showed wide range expansions and repeated recolonizations, including north- and southward dispersals, recolonizations and extinctions. Our results support that the ancestor of South-east Asiatic Herbertus had a Gondwanan origin and arrived in Asia via Indian Plate migration. The uplift of high mountains must have had a strong influence in the diversification and dispersal of the genus. Our results further suggest that climate changes must have had a profound effect on the evolution and biogeography of the species of Herbertus in the northern hemisphere, and might also have influenced the reproductive strategies of the genus. Few genetic differentiations amongst currently recognized species H. aduncus, H. dicranus, H. hutchinsiae, H. stramineus, H. delavayi, and H. kurzii, and amongst H. sendtneri, H. armitanus, and H. circinatus were shown, suggesting that the morphological characters that are currently used for delimiting species should be re-evaluated.