In 1964, Graeme Caughley published two papers on the activity, density, dispersion and social organisation of kangaroos. He made direct observations of the behaviour of two species, and used measures of their faecal pellets to infer habitat use. This work paved the way for many studies of the behaviour of individual kangaroos, the associations between them, and the dispersion of individuals within a population. At an individual level, activity budgets of kangaroos are now known to be influenced by extrinsic factors such as weather, season, and forage conditions, as well as intrinsic factors such as sex and reproductive status. Habitat selection is understood in terms of the requirement for both food and horizontal cover, and past predation continues to play a role in the use of cover, as Caughley first proposed. At the level of associations among individuals, group size is positively related to population density, as Caughley predicted, but is also influenced by habitat structure. Contrary to his views, the high rate of group flux does not represent random movement of individuals. Instead, there are subtle yet persistent associations between related individuals. There is also strong dissociation between population classes when sexual segregation occurs during the non-mating period in seasonally breeding populations. Despite these effects, kangaroos are only weakly constrained by social factors, suggesting that population dispersion may conform to ideal free distribution when food resources are limited. At a management level, this suggests that kangaroos have the potential to redistribute rapidly after culling programs aimed at reducing density in overabundant populations.