In intensive pig husbandry, young pigs are typically housed in an extremely barren environment. Such an impoverished environment may have immediate negative effects on pig welfare (De Jonge et al., 1996; O’Connell and Beattie, 1999). It has been demonstrated that piglets in barren environments engage in more manipulative behaviour, such as biting, nosing and massaging of their pen mates compared to piglets in larger pens with straw bedding (Lammers and Schouten, 1985; Beattie et al., 2000; De Jong et al., 1998; O’Connell and Beattie, 1999). Furthermore, poor rearing conditions may disturb the development of important social skills, as it has been shown that piglets housed in such an environment behave more aggressively than pigs housed in an enriched environment (De Jonge et al., 1996; O’Connell and Beattie, 1999). Most of the researchers who have compared the influence of housing systems have focused on the period after weaning or during fattening (Beattie et al., 2000; De Jong et al., 2000; De Jonge et al., 1996; O’Connell et al., 2004; Olsson et al., 1999). Very few researchers have assessed the influence of pre-weaning housing on the behaviour of the piglets. Pre-weaning housing systems may not only affect the aggression that occurs immediately following the mixing of litters at weaning, but also the agonistic behaviour that occurs in the long term. If the social skills necessary for conflict resolution are underdeveloped in piglets from a poor environment, this will probably result in more intense aggression in certain situations, such as competition for food. The aim of this study was to assess whether three housing systems during the suckling period influence three aspects of piglets’ social behaviour: (i) the play behaviour of piglets before weaning, (ii) the agonistic behaviour caused by mixing of litters after weaning and (iii) the food related agonistic behaviour of pigs during the fattening period. Three pre-weaning housing systems were compared: poor crate (standard farrowing crate without straw), enriched crate (crate with straw, 20% larger), and, as a control, a farrowing pen (pen with straw, 60% larger than the poor crate). At the age of 1, 2 and 4 weeks, play behaviour (locomotor and social play) was recorded for 3h. Four weeks postpartum, piglets were weaned and two litters from the same housing system were mixed in a new pen, with the stocking density corresponding to density in the pre-weaning poor crate. During mixing, the frequency of agonistic behaviour and the proportion of abnormally ended fights (chasing and biting to the hindquarters of the opponent) were recorded for 3h, and after 6h, the number of wounds was counted. Four focal pigs from each litter were chosen (two female and two male) and their agonistic interactions were measured during a food competition test at the age of 3 and 6 months. Our results suggest that the enrichment of the pre-weaning environment through straw and enlarged space probably stimulate pre-weaning play behaviour and substantially reduces the tendency of pigs to behave aggressively during food competition later in life. However, aggression during post-weaning mixing of piglets does not seem to be affected by pre-weaning enrichment.