As the procedures of evaluation research have developed during recent years, it has become clear that the results of evaluation studies have failed to provide policy decision makers with useful information. An approach towards designing and implementing more useful evaluations--the stakeholder survey--is discussed. Stakeholders are individuals, or groups, who have a direct interest in the program being evaluated. The central premise of this paper is that the information needs of stakeholders should be incorporated directly into the formulation of evaluation objectives, and the design of the evaluation to achieve those objectives. A rationale for the stakeholder survey is presented, and its three primary functions (identification, contact and information synthesis) are described, as are application of the approach in a multi-year evaluation of a national supplemental food program. Also considered are key issues involved in applying the approach. The evidence from this study suggests that such surveys are not only feasible, but also of considerable value to the evaluation team. (Author/AL)