Behavioural style and group cohesiveness were tested as sources of minority influence under conditions in which rejection of the minority from the group was possible and under conditions in which it was not. Female subjects (N = 120) were led to believe that they were interacting as a group and that they held a majority position on a relevant issue. The influence agent, ostensibly one of the group members, advocated a minority position throughout their interaction. Three variables were manipulated: group cohesiveness (high or low), behavioural style of the deviate (high or low consistency) and opportunity for rejection of the deviate from the group (possible or not possible). It was predicted that the deviate would be more influential under high cohesive than under low cohesive conditions and that she would be most influential when she was highly consistent and there was no opportunity to reject her. Although both hypotheses were confirmed, unexpected minority influence effects were also found.