The Criança Feliz (Happy Child) parenting program is a home visiting intervention serving over a million low-income Brazilian families. The current study assessed its longitudinal associations with maternal and child speech and gestures in a subsample (n = 145) from the program's impact evaluation research. We performed a between-participants pragmatic randomized trial. Mothers (79.7% Black) and children (17-25 months of age
50.3% female) were randomly assigned to control (n = 73) and intervention (n = 72) groups. One year later, mother-child pairs were recorded playing at home, and the videos were transcribed and analyzed for speech (quantity, vocabulary diversity, sentence complexity, and number of questions) and gestures (quantity). Intention-to-treat analysis revealed that the intervention was associated with the mother's vocabulary diversity (d = .35) regardless of schooling level. No direct or indirect (via maternal speech) associations were observed with the child's speech or gestures. This preliminary evidence suggests that participation in the Criança Feliz program is related to mothers' vocabulary diversity but not to children's speech or gesture measures. The program could focus on maternal gestures, questions, and book reading (complexity feature). Future research evaluating the impact of the program on language outcomes should include pretesting and child follow-up assessments.