It has not been established that a major cause of extinction in birds or any other taxa is failure of metapopulation dynamics: the collapse of a network of ephemeral but discrete populations as movement between them becomes increasingly infrequent. The few data on who goes where and who mates with whom suggest that most species are structured as either a single large population or a small set of source populations and a larger set of sinks. The extinction of the latter is irrelevant to the persistence of the species. However, regional decline of a species in the face of habitat destruction and fragmentation can mimic a failure of metapopulation dynamics, because distinct aggregations of individuals will disappear much as they would if populations in an interacting network were eliminated one by one. Any species with highly restricted range is at great risk of extinction from spatially localized forces, such as cyclones or deforestation. Restricted range rather than inherent weakness is the main reason that so many island species have gone extinct or are endangered. Species with small populations in contact with much larger heterospecific ones with which they are interfertile are threatened with extinction by hybridization. Finally, the disappearance of a species from a site may be due to subtle habitat change, even if this observation seems superficially consistent with some general population theory, such as the dynamic equilibrium theory of island biogeography. Current theory is an inadequate substitute for intensive field studies as a means to address the conservation problems of individual species.