Abstract: Objective: To describe the epidemiology of alcoholism in ED patients. Methods: Over a two-month period, every adult patient brought by ambulance to the ED of a large municipal hospital was prospectively enrolled by questionnaire. Data collected included demographics, previous ED use, triage complaint-related diagnoses, hospital admission rates, and ethanol levels (if determined). The CAGE alcoholism questions were administered to all patients by trained assistants. The only exclusion criterion was the inability to communicate while in the ED. A chisquare analysis was used to compare categorical variables. Results: A total of 2,658 patients were enrolled in the study; 226 were unable to respond to the CAGE questions. Five hundred eighty-eight of the remaining 2,432 patients (24%) were defined as being alcoholic by an affirmative response to at least two of the CAGE questions. All four questions were answered affirmatively by 17% of the total patients. Alcoholic patients were more likely to be male (88% vs 60%), unemployed (87% vs 71%), undomiciled (46% vs 20%), polysubstance users (52% vs 25%), and tobacco users (77% vs 41%), and to have had an ED visit in the previous six months (51% vs 35%) (p < 0.001 for all tests). Ethanol levels ranged from zero to 573 mg/dL. Whereas no positive response to a single CAGE question was predictive of a final diagnosis of alcoholism, a blood ethanol level more than 300 mg/dL predicted an affirmative response to at least two CAGE questions in 97% of cases. Conclusions: Alcoholism should be presumed to be present in a substantial number of patients who present to urban EDs by ambulance.