NINE APPROACHES TO PATIENT OUTCOME EVALUATION ARE OUTLINED: (1) CLIENT COMMENTS ON BENEFITS RECEIVED FROM A PARTICULAR PROGRAM; (2) CLIENT SELF-RATINGS ON PSYCHOPATHOLOGY; (3) ACCOUNTABILITY AUDITS, INCLUDING BIOMETRIC REPORTS SUCH AS DATA ON PATIENT MOVEMENT INTO OR OUT OF TREATMENT; (4) SOCIAL EXPERIMENTS CHARACTERIZED BY THE RANDOM ASSIGNMENT OF CLIENTS TO ALTERNATE TREATMENTS OR TO CONTROL AND EXPERIMENTAL GROUPS; (5) SURVEY OF RELEASED PATIENTS OR THEIR FAMILIES OR GENERAL POPULATION GROUPS FOR THE PURPOSE OF DETERMINING THE EFFECT OF A PROGRAM ON SPECIFIC PERSONS OR ON PART OR ALL OF A COMMUNITY OR CATCHMENT AREA; (6) CASE STUDIES; (7) CLINICAL EVALUATIONS; (8) COST-ANALYTIC TECNIQUES; AND (9) COST-OUTCOME ANALYSIS. GENERAL CONSIDERATIONS INVOLVED IN THE FORMULATION OF PROGRAM OBJECTIVES TO BE EVALUATED ARE WHAT IS TO BE CHANGED, WHO IS THE TARGET POPULATION, WHAT IS THE DESIRED CHANGE; WHETHER A PROGRAM HAS ONE OR MORE OBJECTIVES, AND WHAT IS THE DESIRED MAGNITUDE OF CHANGE. A COMMON DENOMINATOR OF MANY STUDIES DEALING WITH PATIENT OR EX-PATIENT OUTCOME IS BEHAVIOR, ALTHOUGH THE WAY BEHAVIOR IS DEFINED VARIES CONSIDERABLY. AREAS OF PARTICULAR INTEREST IN ASSESSING THE EFFICACY OF DEINSTITUTIONALIZATION AND COMMUNITY-BASED TREATMENT INCLUDE WORK PERFORMANCE, SOCIAL PARTICIPATION, PERCEIVED NORMALITY OF BEHAVIOR, BURDEN ON THE COMMUNITY, APPARENT PSYCHOPATHOLOGY, PERSONAL INTEGRATION, GLOBAL MEASURES OF HAPPINESS, AND CLINICAL RATINGS OF PSYCHOPATHOLOGY. IN THE EVALUATION OF PATIENT OUTCOME, CRITERIA FOR PROGRAM EVALUATION ARE PARTICULARLY RELEVANT. THE CHOICE OF OUTCOME INDICATORS, REALISTICALLY RELATED TO A PROGRAM AND ITS SPECIFIC CLIENTELE, IS ALSO IMPORTANT. REFERENCES ARE INCLUDED. (DEP)