Colombia hosts the largest number of refugees and migrants fleeing the humanitarian emergency in Venezuela, many of whom experience high levels of displacement-related trauma and adversity. Yet, Colombian mental health services do not meet the needs of this population. Scalable, task-sharing interventions, such as Group Problem Management Plus (Group PM+), have the potential to bridge this gap by utilizing lay workers to provide the intervention. However, the current literature lacks a comprehensive understanding of how and for whom Group PM+ is most effective. This mixed methods study utilized data from a randomized effectiveness-implementation trial to examine the mediators and moderators of Group PM+ on mental health outcomes. One hundred twenty-eight migrant and refugee women in northern Colombia participated in Group PM+ delivered by trained community members. Patterns in moderation effects showed that participants in more stable, less marginalized positions improved the most. Results from linear regression models showed that Group PM+-related skill acquisition was not a significant mediator of the association between session attendance and mental health outcomes. Participants and facilitators reported additional possible mediators and community-level moderators that warrant future research. Further studies are needed to examine mediators and moderators contributing to the effectiveness of task-shared, scalable, psychological interventions in diverse contexts.