THERE APPEARS TO BE A CONSENSUS IN EVALUATION LITERATURE THAT THE IMPACT OF EVALUATION RESEARCH ON PROGRAM DECISIONMAKING IS NOT SUBSTANTIAL. UNDERUTILIZATION OF RESEARCH FINDINGS IS A MAJOR PROBLEM. TO ANALYZE VARIABLES AFFECTING THE UTILIZATION OF EVALUATION RESEARCH, 20 FEDERAL HEALTH EVALUATIONS WERE FOLLOWED. THEY INVOLVED FOUR EVALUATIONS OF COMMUNITY MENTAL HEALTH CENTER PROGRAM ACTIVITIES, FOUR HEALTH TRAINING PROGRAMS, TWO NATIONAL ASSESSMENTS OF LABORATORY PROFICIENCY, TWO EVALUATIONS OF NEIGHBORHOOD HEALTH CENTER PROGRAMS, STUDIES OF TWO HEALTH SERVICE DELIVERY SYSTEM PROGRAMS, A TRAINING PROGRAM ON ALCOHOLISM, A HEALTH REGULATORY PROGRAM, A FEDERAL LOAN FORGIVENESS PROGRAM, A TRAINING WORKSHOP EVALUATION, AND TWO EVALUATIONS OF SPECIALIZED HEALTH FACILITIES. KEY INFORMANTS WERE INTERVIEWED TO EXPLORE THE NATURE AND DEGREE OF FEDERAL EVALUATION RESEARCH UTILIZATION. TWO TYPES OF INTERVIEWS WERE CONDUCTED, ONE TYPE WITH GOVERNMENT DECISIONMAKERS AND ONE TYPE FOR EVALUATORS. THE PREDOMINANT IMAGE OF NONUTILIZATION CHARACTERIZING MANY OF THE EVALUATIONS WAS ATTRIBUTED, AT LEAST IN PART, TO A DEFINITION OF UTILIZATION TOO NARROW IN ITS EMPHASIS ON SEEING IMMEDIATE, DIRECT, AND CONCRETE IMPACTS ON PROGRAM DECISIONS. FACTORS AFFECTING THE UTILIZATION OF EVALUATION RESEARCH WERE ORGANIZATIONAL CHARACTERISTICS, ACTORS IN THE SYSTEM (EVALUATORS AND DECISIONMAKERS), AND EVALUATION CHARACTERISTICS (METHODOLOGICAL QUALITY, RELATION TO ORIGINAL OBJECTIVES, BIAS, COMPLEXITY, DEGREE OF DISSEMINATION, RELEVANCE, REPORT FORMAT, POSITIVE VERSUS NEGATIVE FINDINGS, SCOPE OF RECOMMENDATIONS FOR CHANGE, AND TIMELINESS). IN ADDITION, POLITICAL AND PERSONAL FACTORS AFFECTED UTILIZATION. IT IS CONCLUDED THAT THE IMPACT OF EVALUATION RESEARCH IS MOST FREQUENTLY EXPERIENCED AS A REDUCTION IN THE UNCERTAINTY FACED BY DECISIONMAKERS AS THEY ATTEMPT TO DEAL WITH THE COMPLEXITY OF PROGRAMMING REALITY. IMPLICATIONS OF THE PERSONAL FACTOR WITH REGARD TO EVALUATION AND THE USE OF EVALUATION RESULTS ARE DISCUSSED. REFERENCES ARE CITED. (DEP)