Chronic social stress in pigs compromises immune function, reduces ADG, increases activity and skin lesions and affects feeding behaviour but to different extents in individuals and contemporary groups housed in the same way. Assessing the animalsâ€™ perception of chronic social stress is difficult. A large inter-individual distance has been suggested to indicate acute stress immediately following regrouping. The current study hypothesised that large inter-individual distances between grower pigs in groups whose composition was stable for 6 weeks would also be predictive of the expression of other traits sensitive to chronic social stress. Group sizes (20 and 80) and feeder space allowances (32.5 and 42.5 mm/pig) were studied in a 2 Ã— 2 factorial design replicated four times to simulate commercially relevant social environments. Inter-individual distances were calculated for individual resting pigs and defined as the mean distance to the nearest resting neighbour (MNND) and mean distance to all other resting group members (MD). Relationships between MNND or MD and potential stress response variates were examined by appropriate linear or generalised linear mixed models. The mean growth rate was poorer in large groups and the mean number of skin lesions was greater in pens with a small feeding space (P < 0.05), suggesting that these treatments may have increased social tension, but the effects on the mean proportion of activity (P < 0.001) and feeding behaviour (P < 0.05) were not in the expected direction. Pigs which spaced at large distances from other group members had high ADG (P < 0.05). However, large distances between pigs were associated with a higher mean level of activity (P < 0.001) and a lower mean frequency of feeding bouts (P < 0.05) and total feeding duration (P < 0.05), as would be predicted if spacing behaviour was indicative of social stress. Furthermore, fewer lesions were found when pigs rested at a large distance from their nearest neighbour on days 3 and 21 post-mixing (P < 0.05) but, on average, maintained close proximity to other group members in general on day 3 (P < 0.05), suggesting that the group as a whole was more tightly clustered. Spacing behaviour, even in the confined conditions of commercial indoor environments, appears to relate to some measures of chronic stress, although the relationship with skin lesions was complex and that with ADG was not in the expected direction. Measuring spatial responses to activation of neuroendocrine stress response pathways or modifying affective state through anxiolytic drug administration could help to validate how spacing behaviour reflects the perception of chronic social stress. Â© 2013 Elsevier B.V. All rights reserved.