Simple SummaryA worldwide trend amongst zoos is to replacesmall, barren enclosures with large, naturalistic ones intended to provide animals with environments which cater to their behavioural and psychological needs. Evidence suggests that naturalistic enclosures are effective but most studies focus on welfare-related behaviour or human perceptions of the enclosures. To date, little attention has been given to how animals use space in naturalistic enclosures. Our study investigated how a group of chimpanzees at the Johannesburg Zoo used space in a naturalistic enclosure by recording behaviour and space use every 5 min for an hour at a time. We found that the chimpanzees showed a preference for locations within the enclosure which coincided with their previous housing and that the chimpanzees form subgroups which conform to the space of their previous housing (i.e., small, barren enclosure). We suggest that the chimpanzees’ perception of space has been altered by their experience of the previous, smaller barren housing and that this limits their space use in the naturalistic enclosure through what appears to be a self-imposed ‘invisible cage’ around subgroups. Exactly how the ‘invisible cage’ works is unclear but our findings have implications for animal welfare, husbandry and broader conservation of endangered species.Background: Appropriate space is considered paramount for good captive animal welfare. There has been a concerted effort by captive institutions, particularly zoos, to provide captive animals with relatively large, naturalistic enclosures which havehad demonstrated welfare benefits for animals. However, post-occupancy assessments of these enclosures tend to focus on short-term welfare-centredbehavioural effects or human perceptions of the enclosures and their effects and seldom consider spaceuse. We examined the space use of a group of eight captive chimpanzees 5 years after large-scale enclosure modification at the Johannesburg Zoo, South Africa. Methods: Instantaneous scan sampling was used to record behaviour and location of each chimpanzee at 5 min intervals in the new enclosure. From these 6.8 h of data, space-use patterns and subgroup (two or more chimpanzees within 10 m of each other) spacing were considered relative to local environmental variables, social conditions and the location and size of the previous smaller enclosures in which they had been kept. Results: Space use was heterogeneous, with some enclosure zones being used more than others, and 97.5% of subgroups restricted their spacing to the dimensions of the previous housing (10 m × 10 m). Conclusions: This pattern was not explained by individual behaviour, time of day, location, available space, weather, temperature or shade availability, inter-individual spacing or subgroup composition. We suggest the learned helplessness phenomenon may explain these observations and discuss the implications for both animal welfare and endangered species conservation.Regardless of the mechanism, we suggest that such effects could be avoided through the provision of large enclosures for captive animals.