In this address I discuss the contribution to ecological understanding of four aspects of ornithological research, based on long-term or large-scale studies. Long-term population studies of birds, involving marked individuals, have provided, for more than 20 species, information on lifetime reproductive success (LRS) ; that is, the total numbers of young raised by individuals during their entire lives. LRS values are useful because they combine two key measures of performance (annual survival and annual breeding success) into a single overall measure, which could give a close approximation to Darwinian fitness. In any one bird population, LRS values have proved to be markedly skewed, with most individuals in a generation producing no young and a few producing many young. Depending on species, some 3-9% of fledglings in one generation produced half the young in the next generation. The application of island biogeography and metapopulation theory to the bird populations of woods and other fragmented habitats has confirmed the widespread occurrence of the species-area relationship, and the importance of the stochastic processes of local extinction and recolonization in influencing species numbers and distributions among fragments. Certain bird species are found only in habitat patches that are two or more orders of magnitude larger than their individual area needs, as judged from territory sizes. By implication, the more that habitats are fragmented, the more likely are species with special habitat or area needs to disappear, and the more the long-term persistence of the remaining avifauna depends on chance processes. If these points hold for birds, which are among the most mobile of organisms, they surely apply to other less mobile taxa. The nation-wide monitoring of bird populations has revealed long-term declines in many bird species in Britain, associated with modern land-use practices. Research on birds has been paramount in revealing pesticide impacts on non-target species. Some bird populations have been affected directly by pesticides reducing their individual reproductive or survival rates, while others have been affected indirectly by pesticides reducing the abundance of their food organisms, plant or animal. Migration is perhaps more developed in birds than in any other animals. It is impressive in terms of the global scale over which it occurs, the distances travelled, the large proportion of species involved, and the enormous numbers of individual participants. It results each year in huge seasonal changes in the distribution of birds over the earth's surface, linked to seasonal changes in food supplies. In some bird species, annual changes in overall population levels were related to conditions on the wintering range, and in other species to conditions on the breeding range. Large-scale migration studies have raised questions on the factors influencing the geographical ranges of birds, and on the relationship between breeding and wintering areas, and between abundance and distribution.