ANIMAL BEHAVIOUR, 1999, 58, 563–586

Using male and female Alderley Park (Wistar-derived) rats housed in single-sex groups in standard laboratory cages, we looked at the effect of group size (one, three, five or eight) on competitive behaviour and time budgeting (initial and longer term), changes in their serum testosterone (males), corticosterone and antibody concentrations, and organ pathology at age 16 weeks, together with the interrelationships between behavioural measures and pathophysiological indices of social stress. Group size had only limited long-term effects on overall time budgeting and did not affect pathophysiological responses, although there were highly significant differences between individuals in replicate cage groups. Pathophysiology within both sexes showed strong and highly specific correlations with a small subset of behaviours suggesting frustrated attempts to escape from cages, including chewing the cage bars. Escape-related behaviour also correlated strongly with one component of competitive behaviour, Aggressive Grooming within both sexes, although Aggressive Grooming correlated with pathophysiological responses only among males. Females generally showed greater escape-related behaviour associated with greater signs of pathophysiology regardless of the level of aggression shown between cagemates. Major differences in intercorrelated behavioural and pathophysiological responses between replicate groups implied that the individual composition of groups rather than their size had the greater impact on the welfare of the rats, especially among females. This may be consistent with adaptive sex differences in their competitive reproductive strategies. The frequency of apparent escape-related behaviours and Aggressive Grooming, particularly when rats are first introduced into their cage groups, may provide a simple assessment of the welfare implications of particular cage groupings. Copyright 1999 The Association for the Study of Animal Behaviour.