A project for studying ways to optimize utilization of evaluation products in public schools is reported. The results indicate that the negative picture of use prevalent in recent literature stems from the unrealistic expectation that local decision-makers will behave in a classically rational manner. Such a view ignores the political settings of real world evaluation use where the non-use of results may constitute a viable and "rational" alternative. Empirical research has suggested that results are used and in a variety of ways. Factors that have not been demonstrated to affect use include the methodological quality of the evaluations and the timeliness of its report. Factors that do seem to affect use include the report source, content, and receiver; political considerations; and the "personal factor," i.e. the presence of at least one person who cares about the evaluation and its results. A collaborative approach to evaluation capitalizes on the factors that have been shown to affect use, although the practicability of such an approach at the local level may make it unfeasible. (Author/GK)