This article focuses on the formal evaluation of large-scale preventive interventions promoting positive mental health in children and adolescents, using examples of conduct problems. The state of the art in program evaluation is discussed based on quantitative evaluations of mental illness prevention programs. The article reviews current thinking in three related areas of evaluation of health improvement programs for children. First, a case is made for the necessity of evaluations of mental health improvement programs, including a discussion of recent trends in evaluation research in general. Second, more focus specifically on prevention research is encouraged. The dual role of evaluation in the prevention context as both testing efficacy and information theory is presented. Finally, a discussion concerning specifics about design and statistical issues raised by this perspective is presented. The article concludes by addressing how the presented approach to evaluation of prevention programs addresses stages in the prevention research cycle: (1) theory development, which allows for further theoretical understanding of illness processes; (2) implementation clarity, which allows for enhancement of effective, large-scale prevention programs; (3) identification of effective programs to ensure maximum impact; (4) more effective disseminated programs to enhance the likelihood of success; and (5) grounds for improvement to assist in the refinement and success of subsequent preventive interventions. Contains approximately 100 references. (SD)