Alcohol use in the elderly is an increasingly important public health problem. Traditional interventions focus on abuse and dependence in younger persons. Alcohol-related risks and problems in older persons, however, may come from the interaction between alcohol and diminished health or medication use. The proposed study is a randomized trial of the effectiveness and cost-effectiveness of an integrated patient provider intervention to prevent harmful (presence of alcohol-related problems such as liver disease or depression) and hazardous alcohol use (risks for problems) in older adults. The intervention will include a tested computerized screening and education system that was developed especially for older adults and their providers, supplemented by a well-established intervention for physicians. The patient component uses the Computerized Alcohol-Related Problems Survey, which results in printed Patient and Physician Reports with classification of the patient as a harmful, hazardous or non-hazardous drinker and reasons for the classification. The Patient Report references a companion educational booklet developed for older adults. The provider component is based on a physician intervention with proven effectiveness. The proposed research design involves randomization of 28 primary care physicians in four clinics and their eligible patients aged 65+ to the intervention vs. """"""""usual care."""""""" Specific aims are to (1) evaluate the comparative effectiveness of a patient and physician educational intervention to prevent harmful and hazardous alcohol use in the elderly, relative to usual care; (2) evaluate the comparative costs of the intervention, relative to usual care; and (3) evaluate the post-effectiveness of the intervention, relative to usual care. Effectiveness measures include whether patients engage in hazardous and harmful drinking and their health-related quality of life; proximal outcomes examined include knowledge and self-efficacy. The cost-effectiveness analysis of Aim 3 will performed only if the Aim 2 analyses show that the intervention group has higher net costs than the control arm; if the intervention is cost-saving with equally good outcomes, or cost-neutral with better outcomes, then the intervention is cost-effective by definition.