Australia has a fairly large number of species and subspecies of birds, mostly passerines, confined to or largely dependent on mangroves, a closed forest which occurs discontinuously as a narrow chain for thousands of kilometres along tropical and subtropical coasts. About twenty-five forms are mangrove specialists compared with about eleven in New Guinea. North-western Australia is especially rich in birds specialized to mangroves and it is suggested that this region has played an important role in their evolution: isolated populations of mesic species in refuges in north-western Australia apparently became dependent on mangroves as areas of rainforest and monsoonal forest shrank during past arid periods. No such mechanism was possible in New Guinea and north-eastern Australia because large tracts of rainforest remained contiguous with mangroves throughout the Pleistocene. The area in which species and subspecies became specialized to mangroves has often become obscured by their subsequent expansion in range. Within the mangrove habitat of Australia and New Guinea, there have been one or two speciations and about twenty-five subspeciations mainly as a consequence of gaps in mangroves at the Eighty Mile Beach, western side of Joseph Bonaparte Gulf, Gulf of Carpentaria, Torres Strait and possibly the Burdekin Gap. A couple of subspeciations have resulted from gaps in range that may have been caused by competitors rather than by hysical barriers. Patchy distributions and abrupt discontinuities in mangrove birds are discussed in terms of the is!and-like distribution of their habitat. exclusion by possible competitors, geographical barriers, floristic complexity of mangroves and dispersal ability.