METHODS OF MEASURING CORRECTIONAL SERVICES OBJECTIVES, METHODS OF COLLECTING DATA FOR SUCH MEASUREMENTS, AND COSTS AND STAFF TIME REQUIRED FOR ASSESSMENTS ARE GIVEN. FOUR TYPES OF DATA COLLECTION ARE NEEDED: (1) ROUTINE RECORDS KEPT BY PRISONS AND PAROLE AGENCIES; (2) RECORDS OF CRIMINAL ACTIVITY FOR RANDOM SAMPLES OF RELEASED OFFENDERS; (3) PHYSICAL AND PSYCHOLOGICAL EXAMINATIONS FOR A RANDOM SAMPLE OF INMATES; AND (4) INSPECTIONS OF PRISON FACILITIES. TO SET UP AN EVALUATION INITIALLY, IT WILL BE NECESSARY TO FIND A TEAM OF PERSONS INTERESTED IN THE TASK WHO HAVE AT LEAST A YEAR TO SPEND ON SETTING UP THE MEASURE. AFTER THE FIRST YEAR EVALUATIONS CAN PROCEED ROUTINELY. IN ORDER TO MEASURE THE SECURITY OF AN INSTITUTION, ONE SHOULD ANALYZE ESCAPES, INCIDENTS OF FAILURE OF INTERNAL SECURITY, CRIMES COMMITTED BY PRISONERS ON WORK-RELEASE PROGRAMS OR OTHER AUTHORIZED ABSENCES, AND TYPES OF SECURITY FAILURE. CROWDING, SANITATION, AND HEALTH CARE OF PRISONERS ARE SUGGESTED MEASURES FOR HUMANENESS OF INSTITUTION. A SERIES OF BEHAVIOR MEASURES IS FOR EVALUATION OF REHABILITATION EFFORTS IN PRISONS AND THE SUCCESS OF PAROLE SUPERVISON. WAYS TO OBTAIN THIS INFORMATION FROM AVAILABLE SOURCES ARE GIVEN. APPENDIXES DISCUSS OTHER AREAS IN WHICH EVALUATION IS NEEDED AND THE EXPERIENCE OF THE NORTH CAROLINA EFFECTIVENESS MEASUREMENT PROJECT.