This paper examines the relationship between education and mortality in a young population of Italian males. In 1981 several cohorts of young men from specific southern towns were unexpectedly exempted from compulsory military service after a major quake hit the region. Comparisons of exempt cohorts from the least damaged towns on the border of the quake region with similar ones from neighbouring non-exempt towns just outside the region show that, by 1991, the cohorts exempted while still in high school display significantly higher graduation rates. The probability of dying over the decade 1991-2001 was also significantly lower. Several robustness checks confirm that the findings do not reflect omitted quake-related confounding factors, such as the ensuing compensatory interventions. Moreover, cohorts exempted soon after high school age do not display higher schooling or lower mortality rates, thus excluding that the main findings reflect direct effects of military service on subsequent mortality rather than a causal effect of schooling. The authors conclude that increasing the proportion of high school graduates by 1 percentage point leads to 0.1-0.2 percentage points lower mortality rates between the ages of 25 and 35.