Project Northland is a community-wide research program funded by the National Institute on Alcoholism and Alcohol Abuse, for a 5-year period (1990-95). The aim of the study is to prevent or delay onset of alcohol use among young adolescents, as well as to reduce use among those who are already drinkers. Twenty communities were recruited in northeastern Minnesota, an area referred to as the Northland, Arrowhead or Iron Range region, and then were randomly assigned to either Education or Delayed Program conditions. The 10 Education school districts have agreed to participate in 3 years of intervention programs in schools, with parents and in the community-at-large. One group of young adolescents, the Class of 1998 (sixth grade students in the 1991-92 school year), form the study cohort. Surveys (1991-94) of the Class of 1998, their parents, community leaders and alcohol merchants are the primary components of the program's evaluation. Many conceptual and methodological questions emerged during the development of the research protocols for Project Northland over the past 2 years. These questions are the impetus for this article. Specifically, the focus on young adolescents and alcohol use was selected, as contrasted with older adolescents or with multiple problem behaviors. The project was designed using a community-wide model that addresses both supply and demand issues, rather than limited to a school-based model. Intervention strategies and evaluation methods were chosen that could address community-level as well as individual-level behavior change, which required the development and application of new technologies. The rationale for these decisions may be useful to others considering community-wide health promotion efforts.