The relative importance of dispersal and vicariance in the diversification of taxa has been much debated. Within butterflies, a few studies published so far have demonstrated vicariant patterns at the global level. We studied the historical biogeography of the genus Junonia (Nymphalidae: Nymphalinae) at the intercontinental level based on a molecular phylogeny. The genus is distributed over all major biogeographical regions of the world except the Palaearctic. We found dispersal to be the dominant process in the diversification of the genus. The genus originated and started diversifying in Africa about 20 Ma and soon after dispersed into Asia possibly through the Arabian Peninsula. From Asia, there were dispersals into Africa and Australasia, all around 5 Ma. The origin of the New World species is ambiguous; the ancestral may have dispersed from Asia via the Beringian Strait or from Africa over the Atlantic, about 3 Ma. We found no evidence for vicariance at the intercontinental scale. We argue that dispersal is as important as vicariance, if not more, in the global diversification of butterflies.