The paper considers conceptual and methodological issues in studying the scope of alcohol's harm to others. Reasons are suggested for the relative neglect of the topic. The approaches in two relevant research traditions are considered: population surveys on alcohol problems, and economic cost of alcohol studies. Ways of conceptualizing and measuring aspects of the drinker's effects on others are considered, in terms of main types of relationship with the other, and in terms of major societal response institutions. The main types of data tend to measure different levels of severity, with population survey data dominated by less severe problems, and response institution data by more severe problems; so both are needed for a three-dimensional view. Research questions for the field and its policy significance are noted.