This study examined the relationship of group cohesion among heterogeneous and homogeneous groups on individual treatment outcome of child-parent relationship therapy (CPRT). CPRT is a filial therapy model that targets the parent-child relationship as a means for preventing or improving child and/or family problems. This study included 30 parents or caregivers from 9 groups which met for 10 sessions. Participants qualified for this study if their groups ended with at least 3 group members and 2 leaders, all pretest and posttest data on their child between the ages of 2-11 was completed, and if they attended at least 6 of the 10 sessions. Correlation coefficients, t-tests, and effect sizes were calculated. Results demonstrated no statistically significant differences between pretests and posttests on the Child Behavior Checklist (CBCL) for all 30 participants; however, differences in measured effect (η2) between children identified with borderline and clinical behavior problems and children with normal behavior problems suggest that CPRT is more effective among children who demonstrate significant behavior problems. Perceived and observed group cohesion measurements demonstrated no significant difference at the individual outcome level. This finding suggests that group cohesion may not be related to individual outcome. Although there was no significant relationship between group cohesion and individual outcome for this study, results of the group measurements regarding engagement and group cohesiveness, coupled with previous studies on CPRT effectiveness, suggest that CPRT should be utilized in homogeneous groups.