This project aimed to determine the preference of pigs for stalls as a function of stall design, home housing condition and the presence of socially dominant or subordinate group members. Three different experiments (1a, 1b and 2) using similar methodology were carried out. The experiments used a T-maze to determine the preference or aversion of sows for stall designs and social conditions. In the first experiment, experiment 1a, we compared the preference of 16 sows from group housing for free access or open-backed stall which provided 3.6 m2 outside of the stall. We also measured heart rate as a physiological response to the choices being offered to the sow. The second experiment, experiment 1b, was performed in a similar manner but we used 16 sows from gestation stalls. We tested sows from different systems because the type of housing that sows are used to being in may strongly influence the type of system they choose. Sows from stalls may like stalls; sows from group pens may like more space. It was hypothesized that freedom of movement and choice are important to sows, therefore, we predicted that sows, from either housing background, would avoid the locked stall and choose the unlocked stall. The second experiment was designed to determine if group-housed sows tested in pairs would demonstrate a preference for free access stalls (FA) over open-backed stalls (O). 24 sows were used. It was hypothesized that social interactions and rank within a dominance hierarchy may affect a sow’s choice of housing method, therefore, we predicted that subordinate sows would show a stronger preference for the FA stall, where they could protect themselves and spend more time in the stalls than dominant sows. ). The results of experiments 1a and 1b experiments demonstrated that regardless of housing background, individual sows displayed a strong preference for some characteristic of the unlocked free access stall which may include access to space and/or the freedom of choice and movement. We found that there was an increased heart rate prior to choosing unlocked stall which could be due to anticipation of entering the free space, especially for sows with previous experience of long-term confinement. It is clear from the first experiment that further testing is needed to determine the motivating factors that cause the unlocked stall to be a strong preference. The results of the second experiment were quite different from the first experiment. Although a significant preference was found for free-access stalls in the first experiment, the presence of a social companion in the group setting changed the results dramatically. In the second experiment there was no clear preference for either housing system. This may be explained by the fact that sows that were tested already had a clear social relationship, they knew who was dominant and subordinate as they were housed together before testing. Sows spent more time in the pen area than in the stalls but when they spent time in the stalls, they spent more time in the free access stall. There were no significant differences in investigating, lying, and fighting between the stall types FA and O stalls. No stall preference was displayed which may be due to the fact that low levels of aggression occurred, minimizing the urgency to resort to a stall for protection. The results suggest that space, movement and choice are important to individual sows but after a period of socialization, the presence of a dominant or subordinate sow has little impact on the use of housing that allowed sows to isolate themselves. It is possible that free-access stalls may provide a clear benefit during early group formation, or if sows are particularly aggressive to one another in groups. However in stable, calm pairs the benefit is less clear.