Relationship education programs (REPs) are an effective way to enhance relationship communication, prevent relational distress, and increase relationship quality. Most REPs are delivered in a group format; however, there is little known about the influence of group processes on outcomes for these programs, such as group members’ attendance. Therefore, the current study applied a dispersion-consensus model to test the impact of attendance at the member and group levels on group members’ REP outcomes. In a sample of 558 lower income, primarily African American participants, we examined whether individual and group attendance rates influenced posttreatment communication patterns and relationship quality. Results indicated that an individual group member’s attendance was significantly and positively related to their posttreatment relationship quality, although this relationship is complex. Specifically, this relationship was stronger in groups with higher levels of attendance as well as groups with more attendance variability. In addition, results indicated that group members reported better posttreatment relationship quality in groups with less variability in members’ attendance. However, we found a significant interaction between attendance consensus and variability, and an individual group member’s posttreatment relationship quality, suggesting that group members report higher levels of relationship quality in groups where the attendance of the group as a whole is lower yet more consistent. No significant relationships were found for group member’s posttreatment communication patterns. Our findings suggest that the rate and variability in the group’s attendance, as well as an individual group member’s own attendance significantly impacts their posttreatment relationship quality in complex ways.