Using data from The Family Life Project (N = 700), this study examined the pivotal behavior model by testing the hypothesis that children’s social engagement at 24 months mediates the relationship between maternal responsiveness at 24 months and children’s language outcomes at 36 months. Maternal responsiveness during parent-child interaction in a jigsaw puzzle task when children were 24 months old served as a predictor, and children’s social engagement during the same (i.e., persistence, enthusiasm, and compliance) at 24 months, which was assessed during the same task served as the mediator. Children’s language comprehension, language expression, and fluency when children were 36 months old served as the dependent variable. Children’s expressive communication at 24 months was controlled for. The results supported our hypothesis. Higher maternal responsiveness was related to children’s social engagement, which in turn facilitated children’s language development. Children’s language outcomes also were directly related to caregivers’ sensitivity and responsiveness to children’s needs, gestures, and expression.