THE POLICE DEPARTMENT OF EDINA, MINNESOTA, AND NORTH STAR RESEARCH AND DEVELOPMENT INSTITUTE OF MINNEAPOLIS, HAVE CONDUCTED A PROGRAM OF RESEARCH ON THE USE OF PROBABILITY THEORY IN THE ASSIGNMENT OF POLICE PATROL AREAS. THE PURPOSE WAS TO MAKE MORE EFFECTIVE USE OF EXISTING PERSONNEL THROUGH THE USE OF MODERN SCIENTIFIC METHODS. THE MEASURE OF EFFECTIVENESS USED WAS RESPONSE TIME-THE TIME IT TAKES FOR A PATROL OFFICER TO GET TO THE POINT OF NEED AFTER RECEIVING THE CALL. DATA PROCESSING TECHNIQUES WERE USED TO CATEGORIZE REQUESTS FOR POLICE IN THE ENTIRE COMMUNITY AND TO DEVELOP A SERIES OF PATROL AREAS THAT WERE UNEQUAL IN SIZE BUT EQUAL IN CRIME POTENTIAL. EQUAL IN CRIME POTENTIAL MEANS THAT, BASED ON HISTORY OF REQUESTS FOR POLICE AND UPDATING FACTORS, THE PROBABILITY OF A POLICE CAR BEING NEEDED IN ANY OF THE PATROL AREAS AT ANY GIVEN TIME IS THE SAME. IF THE PATROL AREAS OR SUBZONES ARE ASSIGNED TO CARS ON A PURELY RANDOM BASIS, PROBABILITY THEORY PREDICTS THAT PATROL CARS WILL BE CLOSER TO THE POINT OF NEED WHEN THEY ARE REQUESTED THAN UNDER ANY OTHER SYSTEM OF PATROL ASSIGNMENT. THE SPECIFIC GOAL OF THE RESEARCH PROGRAM WAS TO DEVELOP A SYSTEM OF RANDOM PATROL THAT WOULD REDUCE THE TIME REQUIRED FOR A POLICE OFFICER TO RESPOND TO A CALL. ANOTHER DESIRED PRODUCT OF THE PROJECT WAS TO DEVELOP A MODEL PATROLLING TECHNIQUE THAT, IN ADDITION TO SERVING AS THE PATROLLING METHOD FOR THIS PROJECT, WOULD ALSO BE SUITABLE FOR USE BY OTHER POLICE DEPARTMENTS. (AUTHOR ABSTRACT)