It is well-known that social network theory can be applied to free-living and captive populations to determine important patterns of association, and provide an understanding of the relevance of such patterns to the individual and group overall. Highly-social species in zoos can suffer from disrupted bonds and related impoverished welfare due to inappropriate social set-ups. The multitude of social species housed in captivity provides the behavioural researcher with an ideal opportunity to investigate what drives specific association patterns between individuals. And to determine if differing captive environments, between institutions, affect the type and patterning of social bonds present. Flamingos are extremely gregarious birds with a potentially flexible social system, occurring in flocks ranging from a few pairs to over a million individuals. The potential for important bonds to form between birds is therefore high, and past work on flamingo courtship behaviour supports the idea that assortment may affect health, welfare and fitness. As one of the world's most predominate captive species, flamingos are an excellent model species to test the effect of captivity on highly-evolved group-level behaviours. Here we present evidence to show that bonds between individual flamingos are stable and that birds actively assort outside of, as well as during, the breeding season. We show that courtship display performance, as well as aggressive characteristics (i.e. dominance) can affect the potential for information flow and connectivity between birds in a flock. We also explain similarities in social structure between zoos, as well as areas of husbandry (enclosure and flock size) that may alter the opportunities that birds have for developing preferential associations. The idea that flamingos can benefit from social support, and have an in-flock position based on physical attributes and behavioural characteristics, suggests we should consider birds' individualities when mixing flocks and managing groups. This research has wider implications for captive breeding and husbandry, demonstrating the useful of network analysis as a tool for managing captive populations.