Among the different approaches to the study of cardiovascular disease prevention are community-based programs. This type of program concerns a whole community and the intervention takes advantage of the existing service structure and community organization. The evaluation assesses the feasibility, effects on risk factor and disease reduction, costs, process, and other consequences associated with the program. Several such programs have recently been launched in the United States and some other countries. The first major community-based control program was the North Karelia project in Finland, started in 1972 and recently evaluated for its first five-year period. This paper discusses the problems in evaluating community-based CVD control programs on the experiences obtained in the North Karelia project.