This study tests the hypothesis that links between contextual risk and children's outcomes are partially explained by differential parenting. Using multi-informant measurement and including up to four children per family (Mage = 3.51, SD = 2.38) in a sample of 397 families, indirect effects (through maternal differential parenting: self-reported and observed) of cumulative contextual risk on four child outcomes were investigated. Cumulative risk was associated with higher levels of differential parenting and, in turn, with higher levels of behavioral problems. Indirect effects were strongest for attentional and social problems but also evident for aggression. The link between differential parenting and outcome was moderated by favoritism, but this was only evident for maternal report and strongest for aggression.