The Diagnostic Interview Schedule (DIS) is a highly structured instrument that enables lay examiners to gather the clinical information necessary to generate psychiatric disorders according to the DSM-III, Feighner, and Research Diagnostic Criteria. It was developed originally as the diagnostic interview for the Epidemiologic Catchment Area (ECA) survey. Because it adheres to DSM-III and can be used by lay interviewers, thus making it practical for studies involving large samples, it has been used for other population surveys in North and South America, Europe, and Asia. This investigation compares the epidemiology of DSM-III-defined alcohol abuse and addiction in DIS-based population surveys cross-nationally (in St Louis, Mo; Edmonton, Canada; Puerto Rico; Taipei City, Taiwan; and South Korea). We found considerable variation in the lifetime prevalence of alcoholism but a similarity in the age of onset, the symptomatic expression, and the associated risk factors. We also found an inverse correlation between the prevalence of alcoholism and the strength of the association of the risk factors we examined. The work described herein demonstrates the utility of consistent definition and method in cross-cultural psychiatric research. The substantive findings have implications for the definition of alcoholism and for a better understanding of genetic and environmental interactions in its etiology.