To date, nearly half of the work supporting the efficacy of gratitude interventions did so by making contrasts with techniques that induce negative affect (e.g., record your daily hassles). Gratitude interventions have shown limited benefits, if any, over control conditions. Thus, there is a need to better understand whether gratitude interventions are beyond a control condition and if there exists a subset of people who benefit. People high in positive affect (PA) may have reached an 'emotional ceiling' and, thus, are less susceptible to experiencing gains in well-being. People lower in PA, however, may need more positive events (like expressing gratitude to a benefactor) to 'catch up' to the positive experiences of their peers. We examined if PA moderated the effects of a gratitude intervention where youth were instructed to write a letter to someone whom they were grateful and deliver it to them in person. Eighty-nine children and adolescents were randomly assigned to the gratitude intervention or a control condition. Findings indicated that youth low in PA in the gratitude condition, compared with youth writing about daily events, reported greater gratitude and PA at post-treatment and greater PA at the 2-month follow-up.