Suggestions for broadening the utilization of evaluation research findings are discussed. Major program changes, based on evaluation, may not become obvious for several years; program actions may contradict findings and still be based on rational decision making processes; evaluation findings may have far-reaching effects which were not recognized and formally accounted for; personal interaction between evaluator and user may help to shape and redirect the evaluation process. In order to take these factors into account, naturalistic research methods, using techniques such as case studies, field investigations, and participant observations, are recommended. The evaluator's approach is one variable of naturalistic research. The evaluator's influence on the utilization of evaluation information may be determined by: role choice; the extent of which user involvement is encouraged; the amount of attention given to the performance on mandated evaluation tasks; rapport established between the evaluator and users; and the extent to which the evaluator facilitated and stimulated the use of information. Evaluation is defined as a dynamic process; the procedures and outcomes of evaluation studies are influenced by multiple factors as well as interactions between them. Five case studies of evaluation utilization are briefly described. (MH)