The social intelligence hypothesis, originally developed for primates to explain their high intelligenceand large relative brain size, assumes that challenges posed by social life in complex societies withmany group members lead to the evolution of advanced cognitive abilities. In birds, pair-bondedspecies have larger brains than non-pair bonded species, indicating that the quality of socialrelationships better predicts social complexity than group size. Ravens are a long-term monogamousand territorial species, renowned for their sophisticated socio-cognitive skills and complex socialrelationships. Notably, during their early years they live in fission-fusion-like non-breeder societies inwhich social relationships could be of particular importance. Here we observed the development ofdominance and affiliative relationships in 12 hand-raised captive ravens, examining the influence ofage, sex and kinship on social interactions. Furthermore, we investigated at which developmental stepa stable hierarchy emerged, whether third-party interventions played a role and how selectively birdsintervened in others’ conflicts. At 4-5 months post-fledging, we found an increase in socio-positivebehaviour and a decrease in aggression, along with the establishment of a linear dominance rankhierarchy. In line with kin selection theory, siblings exhibited a greater degree of tolerance andengaged in more socio-positive behaviour. In their first few months, ravens frequently intervened inothers’ conflicts but supported mainly the aggressor; later on, their support became more selectivetowards kin and close social partners. These findings indicate that ravens engage in sophisticatedsocial behaviours and form stable relationships already in their first year of life.