Alcohol excise taxation increases prices and reduces per capita consumption (Cook and George Tauchen, 1982; Christopher J. Ruhm, 1995; Douglas J. Young and Agnieszka Bielinska-Kwapisz, 2003). In principle, a taxinduced reduction in per capita consumption of alcohol may be the result of both a reduction in the prevalence of alcohol abuse and the prevalence of moderate drinking, with opposite effects on mortality rates. The net effect on mortality could be either positive or negative and has not been established empirically. Some specific mechanisms by which drinking creates health risks and benefits are well documented. For all age groups, episodes of alcohol intoxication cause death from overdose or from injury resulting from accident or intentional violence (Ralph Hingson and Michael Winter, 2003). Chronic heavy drinking may cause death due to organ damage, including liver cirrhosis (Jurgen Rehm et al., 2003). On the other hand, chronic drinking confers some health benefits on middle-aged people, defined here as persons aged 35–69. Alcohol acts as an anti-cholesterol drug, and epidemiological evidence suggests that moderate drinking is associated with reduced mortality from heart disease and stroke (Giovanni Corrao et al., 2000). Thus an increase in alcohol excise taxes may reduce mortality rates to the extent that it induces a lower incidence of risky drinking and lower prevalence of chronic heavy drinking. But if older people drink too little in response to higher prices, then the result may be increased cardiovascular death rates. In what follows, we combine new estimates of the effect of per capita alcohol consumption on drinking patterns with a summary estimate from the epidemiology literature of relative risks associated with different levels of drinking. We calculate that a permanent reduction of 1 percent in alcohol consumption per capita, induced by a tax increase or some other mechanism, would have little net effect on mortality in middle age. Our sensitivity experiments suggest that the effect may be positive or negative but is always close to zero. Since there is no health benefit from drinking for younger people, and considerable risks, we conclude that the public-health case for increased alcohol taxation is strong.