This paper reviews emergency room (ER) studies from a number of countries which have focused on the association of alcohol and casualties. The review emphasizes studies which used probability sample of patients to represent the population of the emergency facility where the data were collected, and which separated injured patients from patients with medical conditions not due to injuries (the 'non-injured'). Reviewed here are studies concerned with: (1) estimated prevalence of positive blood alcohol at the time of the ER visit; (2) self-reported alcohol consumption prior to the event resulting in a need for ER treatment; (3) patients' descriptions of their usual drinking patterns and alcohol-related problems; (4) predictions of casualties and of alcohol-related casualties. Comparisons of findings from several countries are also presented. Comments on limitations of ER studies as well as other issues pertaining to the usefulness and interpretation of such data, and future directions for research in emergency room populations are discussed.