Background: Nutrition in utero and infancy may causally affect health and mortality at old ages. Until now, very few studies have demonstrated long-run effects on survival of early life nutrition, mainly because of data limitations and confounding issues. Methods: This paper investigates whether exposure to nutritional shocks in early life negatively affects longevity at older ages, using unique individual data and exploiting the exogenous variation implied by natural experiments. In particular, early nutritional conditions are instrumented by exposure to the potato famine of unprecedented severity that the Dutch faced in 1846-47. The individual data are from the Historical Sample of the Netherlands and are augmented by food price data and macro-economic data. The sample used in the study covers lifetimes of 398 individuals exposed and 1,342 individuals not exposed to severe famine during gestation and/or till age three. We compare non-parametrically the total and residual lifetimes of treated and controls per gender. We also estimate survival models in which we control for other individual characteristics and additional (early life) determinants of mortality. Results: Men exposed to severe famine during pregnancy (at least four months) and directly after birth have a significant lower residual life expectancy at age 50 than others, but not at earlier ages. We could not demonstrate any long-run effects for men exposed at ages 0-2 and for women. Conclusion: To our knowledge, this is the first evidence suggesting long-run effects of early nutritional stresses on mortality at old ages for men.