BACKGROUND: In high-income countries, maternal postnatal depression is associated with adverse outcomes in the child. However, few studies have investigated this relation in countries of low and middle income. Furthermore, to our knowledge, no studies have followed up cohorts into later childhood. We aimed to investigate whether maternal depression 6 months after birth is associated with psychological difficulties in a socioeconomically disadvantaged South African cohort of children at age 10 years. METHODS: Birth to Twenty is a prospective, longitudinal, birth-cohort study based in the Soweto area of Johannesburg, South Africa. Mothers and children in this cohort have been followed up at timepoints ranging from before birth to age 10 years. Maternal mood was measured at 6 months with the Pitt depression inventory and at 10 years with the Centre for Epidemiologic Studies depression scale (CES-D). Child psychological functioning was assessed at 10 years with the South African child assessment schedule (SACAS). Our primary outcome was psychological development of children at age 10 years, measured by total score on the SACAS. Secondary outcomes were scores on externalising and internalising subscales of the SACAS. We used t tests to compare psychological outcomes between children whose mother had postnatal depression at 6 months and those whose mother did not have postnatal depression. We examined associations between maternal postnatal depression and child psychological outcomes by multivariate linear-regression analysis, adjusting for socioeconomic status and maternal depression at 10 years, and we used logistic regression to provide odds ratios for associations identified by linear regression. FINDINGS: 1866 mothers completed the Pitt depression inventory 6 months after the birth of their child; of these, 453 (24%) had symptoms of postnatal depression. At the 10-year assessment, 1012 mothers completed the CES-D questionnaire, of whom 747 (74%) were judged to have depression. Sociodemographic characteristics did not differ between mothers with and without depression at both 6 months and 10 years. After adjusting for socioeconomic status and maternal depression at 10 years, children whose mothers had postnatal depression at 6 months were more than twice as likely to have significant psychological difficulties 10 years later compared with children whose mothers did not have postnatal depression at 6 months (adjusted odds ratio 2·26, 95% CI 1·23-4·16). INTERPRETATION: Maternal postnatal depression is associated with adverse psychological outcomes in children up to 10 years later in countries of low and middle income. In view of the increased prevalence of postnatal depression in these settings, this finding has important implications for policy and interventions for children and their mothers. FUNDING: Wellcome Trust (UK), Medical Research Council of South Africa, Human Science Research Council (South Africa), University of the Witwatersrand.