MEASURES OF PATROL OUTCOME, PARTICULARLY PRODUCTIVITY MEASURES, TRADITIONALLY HAVE BEEN USED TO ASSESS PATROL OFFICERS' PERFORMANCE. OTHER TRADITIONAL MEASURES INCLUDE RATING SCALES AND TEST SCORES. THE ACTUAL OBSERVATION OF PATROL PERFORMANCE IS LESS COMMON, AS IS EVALUATION BY CLIENTS OF POLICE SERVICE. IN ADDITION, ARREST QUALITY MEASUREMENT HAS EMERGED IN COUNTERPOINT TO ARREST PRODUCTIVITY COUNTS. IT BECAME APPARENT TO RESEARCHERS INVOLVED IN EVALUATING THE PERFORMANCE OF FEMALE POLICE OFFICERS IN THE NEW YORK CITY POLICE DEPARTMENT THAT, ALTHOUGH MANY EXISTING PATROL PERFORMANCE MEASURES WERE APPROPRIATE IN A STUDY OF POLICEWOMEN ON PATROL, THE MEASURES NEEDED TO BE REFINED AND EXPANDED. THE RESEARCHERS FOCUSED ON FIRST-HAND OBSERVATION AS A MEANS OF OBTAINING INFORMATION ON HOW MEN AND WOMEN CONDUCT THE PATROL PROCESS. A NEW PATROL MEASURE -- CONTROLLING -- WAS DEVELOPED. CONTROLLING BEHAVIOR INVOLVES THE OFFICER'S ATTEMPTS TO INFLUENCE CITIZENS TO TAKE A PARTICULAR ACTION. OTHER MODES OF POLICING PROCESS MEASUREMENT DEVELOPED INCLUDE CITIZEN INTERVIEWS AND REVIEWS OF DEPARTMENTAL ARREST REPORTS. THE NEW YORK STUDY DERIVED THE MOST VALUABLE INDICATORS OF PATROL PERFORMANCE FROM OBSERVATIONS RECORDED BY OBSERVERS WHO ACCOMPANIED OFFICERS ON 8-HOUR TOURS OF DUTY. IT WAS CONCLUDED THAT FIRST-HAND PATROL OBSERVATION, SUPPLEMENTED BY TESTS OF PERFORMANCE IN SIMULATIONS OF PHYSICALLY DEMANDING PATROL EVENTS, IS THE EVALUATION METHOD OF CHOICE. DETAILS OF THE OBSERVATION METHODS DEVELOPED, A COPY OF THE FORM AND CODES USED BY OBSERVERS, AND A COPY OF A SELF-EVALUATION QUESTIONNAIRE FOR PATROL OFFICERS ARE INCLUDED.