--In previous studies, isolated mice are found to be extremely more aggressive than group-reared mice. In most of those studies, however, the pairings of the groupreared mice were randomly determined, without attention to individuals' ranks in their home cages. In this study, the effect of isolation was analyzed by comparing isolated mice with dominants of group-reared mice. In experiment 1 the group-reared mice were randomly chosen as in previous studies. Isolated mice were significantly more aggressive, but the pairings of the group-reared mice were found to be all dominant vs. subordinate. In experiment 2, the group-reared dominants encountered unfamiliar dominants in an unfamiliar place. Dominant mice were found to be less aggressive than the isolated mice. This result was inconsistent with the "territorial dominant hypothesis" on the hyperaggressiveness of the isolated mice. Sequential analysis of behavior clarified that the isolated mice had a sequential pattern quite different from the group-reared mice, and provided an illustrative figure for the "escalation hypothesis". The lack of aggression in the group-reared mice, however, left a question about the sequence similarity of aggressive behavior in isolated and group-reared mice.