The breeding ecology, social organization and communicatory behaviour of the brown creeper (Finschia novaeseelandiae) were studied. Brown creepers lay large eggs, young develop slowly and incubation and nestling periods are long. Mortality in adults is low but high in juveniles. I suggest that the species' breeding strategy is an adaptation to the mild coastal climate of New Zealand. Members of breeding pairs remain together and defend territories throughout the winter. Juveniles form sibling flocks and this association of close relatives may help explain winter mobbing behaviour in the species. The communicatory repertoire consists of 19 vocal and seven visual displays each of which encode several messages. Graded vocal sequences are also used; conceivably because most interactions between conspecifics take place at close range. Three stages in song ontogeny are described. Young males learn song from neighbours they interact with, not fathers. Eight song dialects were studied. The historical model best explains the origin of brown creeper dialects, whereas Payne's (1981) social adaptation hypothesis accounts for the formation and adaptive significance of subdialects. When local and foreign dialects differed markedly, males responded more strongly to the local theme. Similar dialects received equal responses. The lower responses to a foreign dialect may be due to its dissimilarity to the local song. Responses were greater to the songs of strangers than to neighbours. Kowhai Bush males neighbouring a male transferred from Lake Rotorua reacted more strongly to song playback from the transferred bird than did non-neighbours. The results suggest associative learning is important in brown creeper neighbour/stranger discrimination. An unusual vocal display (unison singing) in which two interacting males temporally overlap their songs is described. A unison singing male changes his singing pattern to that of another; such adjustments appear to allow males to define territorial boundaries.