The relationship between alcohol and homicide in America is explored both historically and theoretically, providing the groundwork for two empirical analyses. The first, a theoretical approach, leads to the development of a selective disinhibition hypothesis, the implications of which are tested in a longitudinal analysis of alcohol availability and homicide in 256 U.S. cities between 1960 and 1980. Alcohol availability was found to significantly increase homicide rates. Availability also interacted with city poverty rates, lack of social bonds, and the age structure to further increase the...