We use the Vietnam War draft avoidance behavior documented by Card et Lemieux (2002) as a quasi-natural experiment to infer causation from education to smoking and find strong evidence that education, whether it be measured in years of completed schooling or in educational attainment categories, reduces the probability of smoking at the time of the interview, more particularly the probability of smoking regularly. Interestingly, however, while we find that more education substantially increases the probability of never smoking, our other main finding is suggestive that increased education has a limited impact on smoking cessation behavior. On the one hand there is little evidence that it helps to increase the probability of not smoking regularly at the time of the interview, conditional on having smoked regularly at any time. However, among former regular smokers, those with more education have significantly shorter smoking careers.