Objective: Findings from a number of cohort studies suggest that children who are exposed to maternal stress during pregnancy have an increased risk of behavioral problems. All of the research assessing this association to date has been conducted in developed countries; yet the majority of the child population, and the majority of the burden of psychopathology, exist in developing nations, where different patterns and levels of stressors may exist. The present study set out to examine whether maternal prenatal stress is associated with an increased risk of subsequent child behavioral problems in a developing country. Method: Participants (n = 953) were from Birth to Twenty, a longitudinal birth cohort study based in Soweto-Johannesburg, South Africa, a socioeconomically disadvantaged urban area. Pregnant women completed questionnaires assessing stressors, and children's behavioral functioning was subsequently assessed when they were aged 2 and 4 years. Results: Children whose mothers had high levels of prenatal stressors did not have an increased risk of behavioral problems at age 2 years, but they did at 4 years (adjusted odds ratio 2.52 [CI = 1.04, 6.09]). Partner conflict and family adversity were the stressors most strongly associated with adverse child outcome. Conclusions: These findings highlight the importance of the psychological health of pregnant women in developing countries, both for the mother and her offspring, and the need for greater research in resource-poor settings. (Contains 4 tables.)