THE PROJECTS EXAMINED ARE THE SPECIAL CRIME ATTACK TEAM IN DENVER, COLORADO, THE CONCENTRATED CRIME PATROL IN CLEVELAND, OHIO AND THE PILOT FOOT PATROL IN ST. LOUIS, MISSOURI. EACH OF THE THREE PROJECTS IS EXAMINED INDIVIDUALLY AND THE OFFICIAL CRIME LEVELS DURING THE TIME PERIOD COVERED BY POLICE PATROL PROJECT OPERATIONS ARE ANALYZED. THIS CRIME LEVEL ANALYSIS IS CONDUCTED USING FOUR TIME SERIES MODELS DEVELOPED AS PART OF THE RESEARCH. THESE MODELS PREDICT CRIME LEVELS FOR THE TREATMENT PERIOD BASED ON PAST CRIME LEVELS IN THE AREA. THESE PREDICTED OR EXPECTED LEVELS ARE THEN COMPARED WITH THE ACTUAL LEVELS OF CRIME OBSERVED DURING PROJECT OPERATIONS. FOR EACH CASE AT LEAST ONE OF THE CRIMES EXAMINED (MURDER, RAPE, AGGRAVATED ASSAULT, ROBBERY, BURGLARY) WAS SIGNIFICANTLY LOWER DURING PROJECT OPERATIONS THAN EXPECTED. IN NO ONE PROJECT WERE ALL FIVE CRIMES LOWER THAN EXPECTED, AND NO ONE CRIME WAS LOWER THAN EXPECTED IN ALL THREE CASES. IN GENERAL THE RESULTS SUGGEST THAT WHILE THERE MAY BE NO UNIFORM RELATIONSHIP BETWEEN OVERT POLICE PATROL ACTIVITY AND OFFICIAL CRIME LEVELS THERE IS EVIDENCE THAT PATROLS IMPLEMENTED IN HIGH CRIME AREAS HAVE BEEN ACCOMPANIED BY CRIME LEVELS WHICH ARE LOWER THAN WOULD HAVE BEEN EXPECTED. (AUTHOR ABSTRACT)