Hypotheses for why animals sexually segregate typically rely on adult traits, such as differences in sexual roles causing differential habitat preferences, or size dimorphism inducing differences in diet or behaviour. However, segregation can occur in juveniles before such roles or size dimorphism is well established. In young humans, leading hypotheses suggest that (1) sexes differ in their activity and the synchronisation of behaviour causes segregation and (2) sexes separate in order to learn and maximise future reproductive roles. We reared pheasants, Phasianus colchicus, from hatching in the absence of adults in a controlled environment. Females aggregated with their own sex from hatching, whereas males initially exhibited random association, but segregation became pronounced with age. The increase in segregation corresponded to an increase in sexual size dimorphism. By standardising habitat availability and diet and by removing predation risk, we could disregard the Predation Risk and the Forage Selection Hypotheses operating at this age. Activity budgets did not differ between the sexes, providing no support for the Behavioural Synchrony or the Activity Budget Hypotheses. Both sexes preferentially approached groups of unfamiliar, same-sex birds in binary choice tests, providing support for the Social Preference Hypothesis. Females may segregate to avoid male aggression. Sexual segregation may become established early in development, especially in precocial species, such as pheasants. A clear understanding of ontogenetic factors is essential to further our understanding of adult assortment patterns. Assortment by sex may not be inherent, but rather emerge as a consequence of social interactions early in life.Significance statementHypotheses pertaining to the force driving sexual segregation typically rely on adult traits, such as size dimorphism or differences in sexual roles. However, in some species, animals segregate as juveniles, so that most hypotheses previously invoked to explain sexual segregation in adults are irrelevant. We reared pheasants, Phasianus colchicus, from hatching and monitored multiple aspects of the chicks’ life history in an effort to determine what causes sexual segregation. Females aggregate with their own sex from hatching, whereas males initially have a more random association, but segregation becomes pronounced as both sexes got older, coinciding with greater sexual dimorphism. We controlled for influences of predation risk and dietary/habitat choice and found that activity budgets did not differ between the sexes. Instead, we found that both sexes preferred their own sex when presented with a binary choice, providing evidence that social preference could drive sexual segregation in pheasants.