An individual participant data meta-analysis was conducted to test pre-registered hypotheses about how the configuration of attachment relationships to mothers and fathers predicts children's language competence. Data from seven studies (published between 1985 and 2014) including 719 children (Mage : 19.84 months; 51% female; 87% White) were included in the linear mixed effects analyses. Mean language competence scores exceeded the population average across children with different attachment configurations. Children with two secure attachment relationships had higher language competence scores compared to those with one or no secure attachment relationships (d = .26). Children with two organized attachment relationships had higher language competence scores compared to those with one organized attachment relationship (d = .23), and this difference was observed in older versus younger children in exploratory analyses. Mother-child and father-child attachment quality did not differentially predict language competence, supporting the comparable importance of attachment to both parents in predicting developmental outcomes.