The purpose of this study was to relate the social organization of a population of red junglefowl, Gallus gallus, to evolution theory. During a 7-year study of lifetime breeding success in an unconfined population of red junglefowl, only a small proportion of the most dominant cocks and hens produced most of the adults of succeeding generations. Variability in progeny number reared to independence was such that the genetically effective number of adults in the study population was only 28% of what it would be if the progeny distribution were random (Poisson). Under an isolation by distance model, the effective breeding size was 47% of the harmonic mean (43) of population size over seven years. Combining these two estimates, the genetically effective breeding size of the population was only about 13% of the census number of adults. These results suggest ample basis for random differentiation of local populations in this species, which combined with natural selection and other factors should help to provide favourable conditions for adaptive evolution as envisaged by Sewall Wright in his shifting balance theory of evolution. Social behaviour often greatly restricts effective breeding size of local populations, and so helps to determine the conditions for adaptive evolution favoured by Wright. Wright's theory is based on application of the principles of animal breeding to evolution, and further evidence for this theory is cited from studies on breeding of domestic fowl. The red junglefowl is the principal and perhaps sole ancestor of the domestic fowl. (C) 1996 The Association for the Study of Animal Behaviour