The study used path analysis to explore the longitudinal relations between maternal depression and infant temperament to generate greater inference about the direction of causality. In addition, the analyses examined whether this relationship is mediated by maternal self-esteem.