The Northwest Regional Educational Laboratory was involved in selecting the "Best Report of an Evaluation" for the 1981 awards competition of Division H of the American Educational Research Association (AERA). During this process two things became apparent--there was no procedure or set of standards by which to select a best report, and the quality of reports submitted was highly variable. A procedure and a set of criteria were devised for rating the reports. These should prove useful for providing local report writers a simple set of guidelines for preparing reports that have a greater chance of aiding decision making, and for standardizing the AERA procedure for rating reports. This paper describes both the rating scale and the characteristics of the reports examined. In general it was concluded that evaluation reports need to be clearer on the purpose and audience for the evaluation, on specific evaluation questions, and on sound and useful suggestions for program improvement. Also, evaluation reports need to contain a short, but useful, summary which covers the questions of major concern to the audience of the evaluation. (Author/PN)