This paper examines the long run education and labor market effects from early-life exposure to the Greek 1941-42 famine. Given the short duration of the famine, we can separately identify the famine effects for cohorts exposed in utero, during infancy and at one year of age. We find that adverse outcomes due to the famine are largest for infants. Further, in our regression analysis we exploit the fact that the famine was more severe in urban than in rural areas. Consistent with our prediction, we find that urban-born cohorts show larger negative impacts on educational outcomes than the rural-born cohorts.