Although the need for evaluation of prevention programs is clear, the implementation of evaluation, particularly for large, complex projects with multiple investigators, poses numerous challenges. Program evaluators in these circumstances do not have the same degree of control as, say, experimental psychologists and must find ways to obtain data while being sensitive to local environmental influences. The purpose of this article is to identify for social work evaluators some of the ways program evaluation can differ from true experiments and to discuss issues that may affect the planning and conduct of evaluation efforts. The Appalachia Community Cancer Network is used as a case example, and illustrations of its approach using formative and process evaluation are presented.