The role of economic growth in reducing child undernutrition remains an open and highly debated question that holds important implications for food security strategies. The empirical evidence has been quite contrasted, primarily in regard to the magnitude of the impacts. Yet, most studies have not (appropriately) accounted for the reverse causality between economic growth and child stunting. Using a dataset of 74 developing countries observed between 1984 and 2014, this paper develops a novel approach accounting for the reverse causal effect of stunting on GDP per capita and finds that the impacts of economic growth are much lower than estimated in most previous studies. A 10% increase in GDP per capita reduces child stunting prevalence by 2.7%. In other words, economic growth is modestly pro-poor. We also estimate that a percentage point increase in child stunting prevalence results in a 0.4% decrease in GDP per capita. A back-of-the-envelope calculation suggests that stunting costs on average about 13.5% of GDP per capita in developing countries.