Dynamic social groups are common in sow management. In these groups new animals are added to a resident group at regular intervals, from weekly to every 3-5 weeks. The newly introduced animals are subject to aggression from the resident animals and other newcomers. A question that arises is to what extent does the proportion of animals that are added to the dynamic group affect the aggression and social interaction of the new sows. This study compared the post-grouping injury levels and social integration of sows when 10, 20, 30 or 40% of a dynamic group was replaced. The dynamic groups were comprised of 40 sows. Replacements took place at three-week intervals. Three days prior to new sows being added to the group, 10-40% of the sows were removed. Replacement animals had been housed in groups of 4 for 5 weeks prior to being added to the dynamic group. This is a common procedure that is used for breeding weaned animals. One to four of these groups of 4 sows were added to the dynamic group, representing 10, 20, 30 and 40% replacement rates. Newly introduced animals were exposed to similar amounts of aggression regardless of the replacement rate. However, as replacement rate increased the proportion of aggression derived from other newly introduced sows increased as well. Injuries after one week were similar for all replacement rates. Sows introduced in the 10% replacement treatment spent more time lying outside the kennels (prime sleeping areas) during the first week than did the newly introduced sows in the higher replacement rate treatments. All introduced sows gradually increased their contact time with resident sows over the first week. It would appear that replacing less than 10% of the sows in a dynamic group is quite stressful for those newcomers. Increasing replacement rate beyond 20% does not further reduce the stress on new sows.