Objective: The authors conduct a within intervention group analysis to test whether caregiver engagement (e.g., participation, homework completion, openness to ideas, apparent satisfaction) in a group-based intervention moderates risk factors for foster child outcomes in a state-supported randomized trial of caregiver parent training. Methods: The intervention is delivered in 16 weekly sessions by trained leaders. Outcomes are pre-post change in problem behaviors and negative placements. Results: Analysis of 337 caregivers nested within 59 parent groups show caregiver engagement moderates number of prior placements on increases in child problem behaviors, and moderates risk of negative placement disruption for Hispanics. Conclusions: Variance in parent group process affects program effectiveness. Implications for practice and increasing effective engagement are discussed. (Contains 4 tables and 2 figures.)