FOR HUMAN SERVICE PROGRAMS IN GENERAL AND MENTAL HEALTH AND MENTAL RETARDATION PROGRAMS IN PARTICULAR, THE ISSUE OF PROGRAM EFFECTIVENESS IS PARTICULARLY RELEVANT. POLITICAL RESPONSE TO CITIZEN GROUPS HAS OCCURRED, AND PROGRAM STATUTES AND REGULATIONS INCREASINGLY REQUIRE PROGRAM EVALUATION AS A CONDITION OF FUNDING. THE RELATIVELY SUDDEN AND STRONG PRESSURE IN THE DIRECTION OF EVALUATION APPEARS TO BE THE RESULT OF AN UNEASY POLITICAL ALLIANCE. ONE GROUP OF PROPONENTS SEEMS TO ARTICULATE A GENUINE DESIRE TO PROMOTE IMPROVEMENT IN THE DELIVERY OF HUMAN SERVICES. ANOTHER GROUP SEEMS TO EMPHASIZE FISCAL INTEGRITY WITH LITTLE IF ANY REGARD FOR HUMAN SERVICES. WITH REGARD TO CONTEMPORARY HUMAN SERVICE OBJECTIVES, THE STRESS IS ON COMPREHENSIVE SERVICES THAT WILL MAXIMIZE TOTAL CLIENT FUNCTIONAL POTENTIAL AND MAINTAIN TOTAL FUNCTIONING AT OPTIMUM LEVELS. MEANINGFUL PROGRAM EVALUATION INVOLVES OUTCOME MEASUREMENTS OF SERVICES RENDERED IN TERMS OF CLIENT FUNCTIONAL CRITERIA. EVALUATION IS LIKELY TO BE MOST EFFECTIVE WHERE EVALUATION COMPONENTS ARE BUILT INTO THE DESIGN OF FUTURE SERVICE ACTIVITIES TO BE PERFORMED IN PLANNED PROGRAMS OR PROGRAM SYSTEMS. EVEN SOPHISTICATED EVALUATION SYSTEMS CANNOT BE EXPECTED TO CAPTURE ALL RELEVANT DATA OR GUARANTEE THE INFALLIBILITY OF BUILT-IN JUDGMENTS. WHEN ESTABLISHING PROGRAM POLICY, RELIANCE ON SYSTEM FINDINGS SHOULD BE TEMPERED WITH INTELLIGENT SKEPTICISM. LEGISLATIVE CAUTION AND NOT HASTE SHOULD ACCOMPANY PROGRAM EVALUATION IN ORDER FOR IT TO BE MEANINGFUL. REFERENCES ARE NOT INCLUDED. (DEP)