Tracking child mortality and identifying its causes received promin ence through the global initiative of the Millennium Development Goals (MDGs), [1] which has continued with the Sustainable Development Goals (SDGs). [2] In South Africa (SA), the 1996 census and the 1998 Demographic and Health Survey recorded a reversal of childhood mortality rates at the beginning of the 1990s, coinciding with an expected increase due to HIV/AIDS. [3] However, since then the paucity of sources of reliable estimates of child mortality and incomplete vital registration have made it difficult for the country to track this key indicator, leaving policymakers with limited information to guide interventions. [4] An investigation into the available data sources for measuring child mortality, [5] as well as a demographic analysis, [6] highlighted substantial improvements in the completeness of registration of child deaths between the mid1990s and mid2000s. While the Rapid Mortality Surveillance System [7] provides empirical estimates of childhood mortality to monitor trends in the level of childhood mortality, extensive misclassification of HIV/ AIDS as a cause of death, [8] together with other challenges with medical certification of cause, [9] make it impossible to use the cause ofdeath information at face value. Recently, Bamford et al. [10] called for improved data on child deaths in SA. In addition, the second SA National Burden of Disease Study (SA NBD) [11] has confronted the data issues and provides estimates of the levels and causes of mortality for all ages for the period 1997 2012. Objectives In this article, we present the trends and causes of childhood mortality in SA and the differences between provinces, demonstrating the value of an NBD study in informing health policy. Methods Full details of the methods of the second SA NBD are outlined elsewhere. [11,12] Briefly, the study used Statistics South Africa (Stats This open-access article is distributed under Creative Commons licence CC-BY-NC 4.0.