SummaryIn north-eastern New South Wales, Varied Sittellas occur in sedentary groups or clans holding weakly-defended territories of 13–20 ha. Average group size at one site where several birds were colour-banded was 5.4, while groups observed elsewhere in the region averaged 4.9. Simple pairs occurred in 20% of cases. However, group size varied over the year, some breeding groups amalgamating into ‘clans’ during the non-breeding season. Density in the region varied from 0.2 to about 0.5 birds ha−1. Sex ratios were skewed towards males in samples from two separate districts. Five distinct vocalisations were recognised. Roosting was communal, one colour-banded group using 13 different trees over 120 nights. The group roosted consistently earlier, and awoke later, than other local small passerine species. The breeding season was long, from August to January and second broods were occasionally attempted. Because five out of eight birds disappeared in September-October, mortality and/or dispersal seemed highest during the early breeding season. Most nests were built on dead branches of Broad-leaved Stringybarks Eucalyptus caliginosa. Most clutches (74%) of the species were of three eggs, the remainder being two. Nest success was low (20%; n = 49). Nesting attempts in two groups failed four times per season, although some nests were abandoned before laying. Large territories, feeding of the incubating female, and long periods of incubation, nestling and juvenile dependency, as well as a specialised foraging niche and cryptic prey all suggest that food may be limiting for this species. Thus, helpers may normally be required to provide young with sufficient food. However, I argue that group-living and philopatry in this species probably developed to increase foraging efficiency and reduce vulnerability to predators. The species offers exceptional scope for studies of the influence of habitat on sociality and cooperative breeding, and the effects of group size on foraging success.