Abstract
A risk function analysis of data is reported from the 2005 National Alcohol Survey on average daily volume of alcohol and frequency of consuming five or more drinks per day with reporting any injury during the preceding year, an injury that was treated, or an injury that was treated in the emergency room (ER). Data suggest the association of alcohol and injury based on ER studies is not representative of all injuries, and should be taken into account in deriving estimates of risk of injury related to alcohol and alcohol-attributable fractions based on ER studies.