A randomized controlled trial was used to examine the impact of an attachment-based, teacher-child, dyadic intervention (Banking Time) to improve children's externalizing behavior. Participants included 183 teachers and 470 preschool children (3-4 years of age). Classrooms were randomly assigned to Banking Time, child time, or business as usual (BAU). Sparse evidence was found for main effects on child behavior. Teachers in Banking Time demonstrated lower negativity and fewer positive interactions with children compared to BAU teachers at post assessment. The impacts of Banking Time and child time on reductions of parent- and teacher-reported externalizing behavior were greater when teachers evidenced higher-quality, classroom-level, teacher-child interactions at baseline. An opposite moderating effect was found for children's positive engagement with teachers. (author abstract)