THE MOST FREQUENTLY HEARD REASON FOR THE DISCOUNTING OF EVALUATION RESEARCH IS POOR METHODOLOGY AND DESIGN. THE LITERATURE ON NONUTILIZATION EMPHASIZES THAT EVALUATORS OFTEN FAIL TO MEASURE THE MOST SIGNIFICANT VARIABLES. FOR EXAMPLE, ORGANIZATIONAL THEORY SUGGESTS THAT ACTUAL GOALS, PROCESSES, AND THE STRUCTURE OF AN ORGANIZATION TEND TO VARY FROM THEIR DESCRIPTIONS. IN THE MENTAL HEALTH FIELD, SELECTED OUTCOME MEASURES ARE OFTEN TOO ABSTRACT TO PROVIDE SATISFACTORY IMPLICATIONS FOR THE MODIFICATION OF INTERVENTION STRATEGIES. A NUMBER OF RESEARCHERS HAVE SUGGESTED POSSIBLE STRATEGIES FOR DEALING WITH THE PROBLEMS OF EVALUATING 'PHANTOM' PROJECTS, SELECTING MEANINGFUL COMPARISON GROUPS, AND IDENTIFYING RELEVANT VARIABLES. WHAT THESE STRATEGIES HAVE IN COMMON IS AN EMPHASIS ON SOLVING PROBLEMS BY CONSULTING, COMMUNICATING, AND GENERALLY BECOMING MORE CONGRUENT WITH ADMINISTRATORS AND PROGRAM STAFF IN DEALING WITH ISSUES. THE STRATEGIC COMBINATION OF ETHNOGRAPHIC AND PARTICIPANT OBSERVATION FIELD METHODS, RECORD AUDITS, AND PERSONAL INTERVIEWS CAN BE USEFUL IN EVALUATING BEHAVIORAL GOALS AND ACTIVITIES OF A PROGRAM. ANOTHER WAY OF IMPROVING THE UTILITY OF EVALUATION RESEARCH IS TO STUDY THE DYNAMICS OF THE PROGRAM PROCESS. THE TRADITIONAL GOAL ATTAINMENT MODEL IN EVALUATING RESEARCH DOES NOT CONSIDER THE IMPORTANT TASK OF IDENTIFYING THE DYNAMICS OF THE INTERVENTION THAT ARE DIFFERENTIALLY RESPONSIBLE FOR PROGRAM FAILURE OR SUCCESS. OTHER REASONS FOR THE NONUTILIZATION OF EVALUATION RESULTS CONCERN COMMUNICATION AND THE INTERESTS OF INVOLVED PARTIES (PROGRAM OFFICIALS, POLICYMAKERS, PROGRAM ADMINISTRATORS, AND EVALUATORS). THE EXTENT TO WHICH EVALUATION RESEARCH IS REDIRECTED FROM THE ASSESSMENT OF OUTCOMES AND GOAL ATTAINMENT TO AN EXCLUSIVE CONCERN WITH PROCESS IS A COMPLEX AND PROBLEMATIC ISSUE. REFERENCES ARE INCLUDED.