THE 11 PILOT SCHOOLS WERE PAIRED WITH 11 CONTROL SCHOOLS IN THIS PORTLAND, OREGON, PROJECT. DURING THE 2-YEAR INSTALLATION TRIAL (FEBRUARY 1, 1973-JANUARY 31, 1975), THE CONTROL SCHOOLS' BURGLARY FREQUENCY DECLINED BY 11 PERCENT, COMPARED TO THE TARGET SCHOOLS' DECLINE BY 27 PERCENT. THE CONTROLS' PROPERTY LOSS INCREASED 37 PERCENT, COMPARED TO A 42 PERCENT DECREASE IN THE ALARM-EQUIPPED SCHOOLS. IN CONTROL SCHOOLS, THE VALUE OF PROPERTY RECOVERED INCREASED BY 27 PERCENT, WHILE IN TARGET SCHOOLS IT DECREASED 61 PERCENT. HOWEVER, THIS REFLECTS THE QUICKER ARRIVAL OF POLICE WHILE THE PROPERTY WAS STILL IN THE BUILDING; THIS REDUCED THE VALUE OF BOTH PROPERTY STOLEN AND RECOVERED. CLEARANCE RATES FOR BOTH PILOT AND CONTROL SCHOOLS INCREASED 6 PERCENT DURING THE TRIAL PERIOD. DURING THE TEST, THE SOCIOECONOMIC CHARACTERISTICS OF THE SCHOOL NEIGHBORHOODS REMAINED ESSENTIALLY THE SAME. EIGHT SOCIOECONOMIC FACTORS WERE STUDIED IN CONNECTION WITH BURGLARY FREQUENCY AND PROPERTY LOSS. ONLY 'PERCENTAGE OF STUDENTS ATTENDING SCHOOL' SHOWED A SIGNIFICANT ASSOCIATION; 'R VALUE' CORRELATION BETWEEN ATTENDANCE AND BURGLARY FREQUENCY WAS -.54, BETWEEN ATTENDANCE AND PROPERTY LOSS -.60. BOTH WERE SIGNIFICANT AT THE .001 LEVEL. THIS FINDING SUPPORTS PREVIOUS EVIDENCE THAT ALIENATION FROM SCHOOL IS HIGHLY ASSOCIATED WITH CRIMINAL BEHAVIOR. A COST-BENEFIT ANALYSIS OF THE ALARM INSTALLATION AND MAINTENANCE VERSUS VALUE OF PROPERTY RECOVERED, STOLEN, OR DAMAGED SHOWED THAT COSTS EXCEEDED SAVINGS BY ABOUT $287,828. HOWEVER, THIS IS BASED ON PREINSTALLATION NET LOSSES. SINCE THE LOSS RATES HAVE BEEN CLIMBING STEADILY, THE ANNUAL SAVINGS OF THE SYSTEM COULD ACTUALLY BE MORE. IF BENEFITS WERE FIGURED ON THE BASIS OF LOSS RATES IN COMPARABLE SCHOOLS RATHER THAN PROJECTED FROM PREINSTALLATION LOSS RATES, THE SYSTEM WOULD PAY FOR ITSELF. TABLES PRESENT DATA COLLECTED FOR THIS STUDY, INCLUDING BURGLARY INCIDENCE BY TIME, BY SCHOOL, AND BY MONTH, AS WELL AS PROPERTY LOSS FIGURES, CLEARANCE RATES, AND OFFENDER CHARACTERISTICS.