Early use of alcohol by youth contributes to initiation of a chain of risky behaviors, and early intervention may prevent a great variety of harmful consequences. In this study, we propose a randomized, controlled design to test the utility in a large, urban Pediatric Emergency Department (PED) of a brief negotiated interview and active referral (the BNI-AR) with patients aged 14-21 who screen positive for risky and dependent drinking in order to reduce alcohol consumption, alcohol related injury and other alcohol related risk behaviors and link young patients with resources and, where appropriate, with the alcohol treatment system. We will screen 10,000 PED patients to detect those with an alcohol related problem. In order to control for assessment reactivity, we will randomize participants to either an intervention group, a control group that receives baseline assessment using a panel of standardized instruments, or a control group that receives no assessment. The intervention consists of normative feedback, review of the pros and cons of current alcohol use, assessment of readiness to change, review of any association between alcohol and injury and/or between alcohol and the patient's presenting medical problem, evaluation of strengths and assets, negotiation of a contract for change, referral to alcohol treatment and/or other resources, and a booster session at 10 days. We follow the three groups at 3 and 12 months to analyze outcomes related to consumption, alcohol problems, injury, fighting, school attendance, safe sex and use of treatment and other resources. We propose to expand in a PED setting with: 1) a large study sample of 1275 14-21 year olds who vary in age, risk profile, racial/ethnic background, reason for visit and severity of alcohol related problems; 2) the use of culturally competent community outreach youth workers to conduct comprehensive screening and motivational intervention, 3) expansion of outcomes to include self report of reduction in intentional and unintentional injury, alcohol-related risks (number of episodes of drinking and driving, physical fights and unsafe sex), and objective measures of motor vehicle record violations, hospital ED visits and trauma admissions, and contact with the alcohol treatment system at one year follow up; 4) analysis of independent effects of age group at entry, age of onset of drinking, quantity and frequency of drinking at baseline, risk-taking propensity and hospital admission; and 5) introduction of a control for assessment reactivity.