THE PRESSURE FOR RELEVANCE EXERTED BY FEDERAL POLICYMAKING OFFICIALS WHOSE AGENCIES SUPPORT SOCIAL RESEARCH HAS TAKEN A VARIETY OF FORMS. FEW PROPOSED SOLUTIONS TO THE RELEVANCE PROBLEM HAVE BEEN BASED ON A CLEAR IDEA OF WHAT CONSTITUTES 'POLICY-RELEVANT' OR 'SOCIALLY USEFUL' RESEARCH. THE MANAGEMENT OF SOCIAL RESEARCH FOR POLICY RELEVANCE, A DIFFICULT TASK UNDER THE BEST CIRCUMSTANCES, IS MADE MORE DIFFICULT BY THE COMPLEXITIES OF POLICYMAKING. THE TIME-CONSUMING, ACTION-FORCING, INCREMENTAL, AND ADAPTIVE NATURE OF THE POLICYMAKING PROCESS HAS IMPORTANT, CONFLICTING IMPLICATIONS FOR SOCIAL RESEARCH. IN SHORT, THE RESEARCH MANAGER'S DILEMMA IS THIS: INVEST IN THE FUTURE BY SUPPORTING BASIC RESEARCH, WHICH USUALLY HAS LITTLE CURRENT INTEREST BUT OFFERS THE POSSIBILITY OF SIGNIFICANT LONG-TERM PAYOFFS; OR MEET IMMEDIATE NEEDS WITH RESEARCH WHICH HAS GREATER SUPPORT BUT MAY NOT WEIGH SO HEAVILY IN THE LONG RUN. THE CHOICE OF CRITERIA FOR MEASURING POLICY RELEVANCE DEPENDS ON ONE'S VALUES AND PERCEPTIONS CONCERNING THE APPROPRIATE FEDERAL ROLE IN SOCIAL RESEARCH. FEDERAL SOCIAL RESEARCH MANAGEMENT HAS BEEN CHARACTERIZED BY TOO LITTLE APPLICATION OF RELEVANCE CRITERIA IN THE PLANNING STAGES AND BY A FOCUS ON INDIVIDUAL PROJECTS RATHER THAN ON MULTIYEAR, MULTIPROJECT PROGRAMS. THE SOLUTION TO THE RELEVANCE PROBLEM HAS TWO ASPECTS: RECOGNITION OF THE COMPLEXITIES OF THE KNOWLEDGE-INTO-POLICY PROCESS AND OF THE MANY CRITERIA THAT ARE APPROPRIATE TO ASSESSING RELEVANCE; AND CONSCIOUS APPLICATION OF CRITERIA OF RELEVANCE DURING THE FORMULATION OF SOCIAL RESEARCH AGENDA. (LKM)