SECURITY SURVEILLANCE SYSTEMS USING CCTV ARE BEING DEPLOYED WITH INCREASING FREQUENCY FOR THE PROTECTION OF PROPERTY AND OTHER VALUABLE ASSETS. PRIMARY BASELINE PERFORMANCE REQUIREMENTS CHOSEN FOR RELEVANT FAILURE ASSIGNMENT IN EVALUATION WERE RESOLUTION OF DETAIL AND DISCRIMINATION OF GRAY SHADES. MINIMUM VALUES OF RESOLUTION AND DISCRIMINATION WERE DETERMINED IN PRIOR PERFORMANCE TESTS WITH SIMULATED INTRUDERS APPROACHING A PERIMETER FENCE AND VARIABLE GLIMPSE TIMES FOR ASSESSMENT. BOTH RESOLUTION AND GRAY DISCRIMINATION REQUIREMENTS WERE DETERMINED FOR DAY AND NIGHT CONDITIONS, AND THESE MINIMUM VALUES BECAME THE BASIC DEFINITION FOR RELEVANT FAILURES. A RESOLUTION CHART CONTAINING HORIZONTAL AND VERTICAL RESOLUTION LINES RELATED TO PARTICULAR SYSTEM REQUIREMENTS, AS WELL AS GRAY SHADES, WAS UTILIZED TO ASSESS RELIABILITY. OTHER MEASURES OF PICTURE QUALITY WERE THE PERCENT OF PICTURE COVERAGE AND THE SIZE, NUMBER, AND SPACING OF CAMERA BLEMISHES. TEST STIMULI CONSISTED OF SIMULATED INTRUSIONS. AFTER AN INTRUSION, NORMAL OPERATOR RESPONSE WAS TAKEN AS IF A REAL INTRUSION HAD BEEN DETECTED. VARIOUS CAMERA SECTORS, ASSOCIATED WITH BLIND ZONES OF SELECTED INTRUSIONS, WERE INPUT TO THE MICROPROCESSOR LOGIC FOR DISPLAY AS APPROPRIATE. ALARMS WERE ALLOWED TO STACK UP, I.E., SOME ALARMS WERE BEING ASSESSED, WAITING TO BE ASSESSED, OR NOTED AS HAVING BEEN PREVIOUSLY ASSESSED. AS ALARMS WERE ENTERED AND ULTIMATELY RESET, EVERY LOGICAL COMBINATION OF ALARM PROCESSING WAS CHECKED. AN ADDITIONAL STIMULUS CONSISTED OF CAMERA FAULTS. POWER WAS ROUTINELY INTERRUPTED TO CAMERAS TO SIMULATE THE LOSS OF VIDEO AND/OR CABLE TAMPERING. THESE ALARMS WERE PROCESSED AS IF THE CAUSE WAS UNKOWN. ALL TEST STIMULI WERE EXERCISED AT LEAST TWICE DAILY TO ASSURE THAT THE SURVEILLANCE SYSTEM WAS OPERATIONAL UNDER VARIOUS EXTERNAL LIGHTING CONDITIONS. IT IS SHOWN THAT THE METHOD OF RELIABILITY DETERMINATION HAS APPLICABILITY TO OTHER TYPES OF IMAGING SURVEILLANCE SYSTEMS BESIDES CCTV. DATA, ILLUSTRATIONS, AND REFERENCES ARE INCLUDED. (DEP)