This paper estimates the eﬀects of maternal stress and malnutrition using the 1959-1961 Chinese famine as a natural experiment. Observed forty years later in the 2000 China Census (1% sample), Famine survivors showed impaired literacy, labor market, wealth, and marriage market outcomes. In addition, maternal malnutrition reduced the sex ratio (males to females) in two generations – those prenatally exposed and their children – presumably through heightened male mortality. This tendency toward female cohorts is interpretable in light of the Trivers-Willard (1973) hypothesis, according to which parents in poor condition should skew the oﬀspring sex ratio toward daughters. Hong Kong Natality micro data from 1984-2004 further conﬁrm this pattern. The persistence of poor nutrition in China – particularly in rural areas and among girls – suggests that health and economic outcomes will be compromised well into the 21st century.