Public health and physical education are both enthusiastic about the role schools might play in reducing childhood obesity. Whether or not this is realistic, these hopes tend not to be based on an assessment of previous attempts to use schools to influence the health-related behaviours of an entire nation. This is a problem because it cuts off one source of insight about the reasons for past successes and failures. This presentation begins with a summary of 100 years of research into the effectiveness (or otherwise) of alcohol education in American public schools and looks for parallels in contemporary research into school-based anti-obesity interventions. In many respects, the obstacles are the same in both cases: insufficient teacher knowledge, grassroots resistance, and the long list of competing priorities that schools must juggle. In a broader sense, though, this presentation considers the enduring logic of the crusade in school-based interventions, a logic that assumes that if the cause is sufficiently just, it must be worth pursuing. One consequence of this is that the rationale for intervention tends to rest, at best, on a largely untested theory or, at worst, emotive statements about the nature of the crises and the risks of inaction.