THE HISTORICAL DEVELOPMENT OF COMMUNITY-BASED CORRECTIONS IS REVIEWED, AND THE INADEQUATE UTILIZATION OF PLANNING TECHNIQUES IN THIS DEVELOPMENT IS NOTED. METHODOLOGICAL CONSIDERATIONS ARE EXPLORED FOR INCREASING THE UTILIZATION OF PROGRAM EVALUATION AS A FEEDBACK MECHANISM FOR PLANNING AND HALFWAY HOUSE DATA ANALYSIS. EVALUATIVE AND MANAGEMENT OPERATIONS OF HALFWAY HOUSES CAN BE SIMPLIFIED BY THE DEVELOPMENT OF MEASURABLE AND REALISTIC GOALS AND BY OUTLINING THE PERCEIVED INTERACTION OF GOALS IN AN OBJECTIVE HIERARCHY. HALFWAY HOUSES ARE EFFECTIVE TREATMENT MODALITIES FOR HELPING EX-OFFENDERS REINTEGRATE INTO THE COMMUNITY. EX-OFFENDERS WHO RECEIVED HALFWAY HOUSE SERVICES SCORED SIGNIFICANTLY HIGHER ON MEASURES OF RELATIVE ADJUSTMENT THAN A COMPARATIVE GROUP WHICH DID NOT RECEIVE THESE SERVICES. EVALUATION RESULTS UTILIZING QUASI-EXPERIMENTAL DESIGNS AND DICHOTOMOUS MEASURES OF OUTCOME BASED ON OFFENSE DISPOSITION ARE SOMETIMES MISLEADING OR INVALID BECAUSE THEY DO NOT CORRECT FOR DIFFERENCES IN EXPERIMENTAL AND COMPARISON GROUPS. DIFFERENCES EXIST IN THE BENEFITS RECEIVED BY RESIDENTS OF HALFWAY HOUSES. CHANGES IN CORRECTIONAL POLICIES CAN HAVE AN EFFECT ON PROJECT LOADS OF CLIENTS REFERRED TO HALFWAY HOUSES. THESE STUDY CONCLUSIONS ARE BASED ON AN EXAMINATION OF HALFWAY HOUSES IN OHIO. PROCEDURAL INFORMATION ON THE STUDY IS APPENDED, AND A SELECTED BIBLIOGRAPHY IS INCLUDED. (DEP)