A problem, endemic to program evaluation, centering around the inability of evaluations to produce meaningful information, is outlined. A general strategy to be utilized in responding to the problem, the ethnographic approach, is presented. The discussion of an ethnographic evaluation model focuses upon: the role of the participant observer, establishing oneself as a participant observer evaluation, and the handling of sensitive information. The exploratory activities out of which came the initial issues to be investigated are described, along with some observations about the process of delimiting the investigation, and of initiating "personal" relationships. Discussion is interspersed throughout the paper, focusing upon the methods used to collect data, and the line of investigation pursued during the study. Finally, the advantages of the ethnographic approach to evaluation as compared to the experimental approach are discussed. (Author)