The study determined that the composition of the treatment group does not make a difference to the overall therapeutic climate in sex offender treatment. Group members’ ratings regarding social climate were in the medium to high range, indicating a generally positive group environment. As a rule, the homogenous groups were not superior to the mixed groups; however, groups did differ significantly in terms of expressiveness. The lack of significant differences between the recidivism rates of the child abusers in the different types of groups provides further evidence that group composition does not need to be a factor in assembling a group that will have a positive therapeutic climate. In a group of inmates, including some sexual offenders, group relationship building and cooperation were reported to be important processes and goals. Group therapy with sexual offenders allowed the members to realize that others had committed similar acts that they also felt bad about. By letting inmates see problems like their own from a safe distance, allows a more objective focus on their own issues. This can be achieved in groups with any mix of sex offender types, not just those who have committed the same type of offense. Therefore, although these results are preliminary and links between group composition, positive group environment, and recidivism remain to be more definitively demonstrated, this study does not support one particular group composition over another. The Group Environment Scale (GES), which measures social climate, was administered to 73 male sexual offenders in groups of those who only victimized adults or children (5 groups) or men who both victimized adults and those who victimized children (5 groups). The results are discussed in the light of mixing child molesters and rapists in group-based CBT. Limitations and future research are detailed. Tables, references