There are two competing views regarding the presence of social fathers on childrens’ cognitive ability: (1) either the social father provides more financial resources which benefit the children or (2) the mother with new partners may shift the focus away from the children. Previous research focused on older children or adolescents and ignored the self-selection problem. We use data from the Fragile Families and Child Wellbeing Study (FFCWS), and a sample of younger children. Using propensity score matching method (nonparametric methods), we find that children with social fathers scored around three points less in a cognitive ability test than children living only with biological mothers (assuming that self-selection is based on observables). The result remains robust when using a control-function analysis (parametric method).