This report presents the results of a yearlong study of the process of evaluation use in a large city school district. Based on extensive naturalistic data-gathering, the research documents how local school administrators use evaluation information generated by the district's research and evaluation unit. The significance of the research is the framework presented for conceptualizing the evaluation use process in a local education agency, including the following concepts: three new or revised use factors ("espoused theories" vs. "theories-in-use," an expanded personal factor, and the clout factor); the types of use (signalling vs. charged use); the evaluation use process; and the domain of charged use. Concluding recommendations are directed toward local school evaluators and users. They emphasize the importance of evaluator sensititity to organizational complexities within school districts, urging that most time and resources should be spent on evaluations supported by key administrators who have internal clout. Evaluation reports required to "signal" that the ongoing program responsibility is met will be conducted to fulfill external requirements, but these reports may not have the potential to directly effect local program change. Finally, implications for future research, including the call for more naturalistic evaluation use studies, are presented. (Author/PN)