The impact of minority dissent on group-level outcomes is explained in the current literature by two opposing mechanisms: first, through cognitive gains due to a profound change induced by minority members in the individual cognitions of the majority members, and second, through socio-affective process losses due to social rejection and relationship conflict. Groups are most effective in information processing if they succeed in solving this opposition and reduce the negative impact of process losses. The present study addresses this opposition using an experimental design in which we crossed minority dissent (presence vs. absence of minority dissent) with change in membership (groups with vs. groups without change in membership) to determine which condition leads to the highest group cognitive complexity. Our results show that groups with a history of dissent and where the deviant left the group have the highest cognitive complexity, followed by groups that experienced dissent and where no change in group membership took place. The groups without a history of dissent have the lowest cognitive complexity.