AbstractAs part of a longitudinal evaluation of a violence prevention program in Medellin, Colombia, researchers used focus groups to explore participants’ perceptions of their experience during the program and to help analyse outcome data 4 years later. A quasi-experimental multi-year evaluation of the Medellin early prevention of aggression program showed statistically significant associations between program participation among children aged 3–9, their families, and teachers, and patterns of prosocial behaviour, non-violent parenting practices, and less severe teacher discipline. Focus group participants (parents, teachers, and children) selected through stratified random sampling were inconsistent in their capacity to recall program characteristics that contributed to intervention fidelity, sustainability, and individual and family outcomes. The difficulties and benefits of employing focus groups to help interpret quantitative longitudinal research findings are discussed.Keywords: prevention of aggressionfocus groupslongitudinal researchsocial program evaluationparents and caregiverseducatorschildren