SUMMARYThe ways in which the numbers of tropical sea‐birds might be limited are considered; it is argued that food is the only factor likely to be generally effective in limiting numbers, but it seems improbable that food shortage could exert a density‐dependent control of the mortality of the birds outside the breeding season.Wynne‐Edwards' hypothesis that colonies of sea‐birds are able to keep their own numbers below the level at which food shortage would become acute, primarily by exerting control on the output of young, is rejected as unproven and improbable.It is suggested that colonies of tropical oceanic birds deplete the food in the waters round them, and that as the populations increase competition for food becomes more intense, and relatively fewer adults succeed in raising chicks. This would provide a density‐dependent control of the output of young and could regulate the numbers of the birds.The peculiarities of long‐lived sea‐birds (e.g. clutches of one, long fledging periods, deferred maturity) which Wynne‐Edwards suggests are adaptations evolved in order to lower the reproductive rate until it balances the mortality, apparently could not be evolved as such; they are more probably adaptations enabling the birds sometimes to raise single chicks in spite of competition for food that makes it impossible to raise more than one. It is considered that variation in the age of first breeding provides an important supplement to variation in reproductive success in regulating the numbers of long‐lived tropical sea‐birds.The possible applications of this hypothesis to sea‐birds breeding in higher latitudes are briefly considered, as are its implications in relation to conservation and exploitation of pop