CHANGES OVER TIME IN CRIME AS MEASURED BY REPORTED CRIME COUNTS AND VICTIMIZATION SURVEY COUNTS WERE COMPARED. A SIMPLE ANALYSIS WAS DONE USING DATA FOR LEAA'S NCS FOR 1973-75; THIS PROVIDED SOME UNDERSTANDING OF WHY UCR RESULTS FOR THESE YEARS SHOWED A MUCH GREATER INCREASE IN CRIME THAN THE LEAA/CENSUS BUREAU NATIONAL CRIME SURVEYS. THE DISPARITY MAY BE DUE, IN PART, TO AN ACTUAL INCREASE IN PART I UCR OFFENSES OR THE TYPE OF CRIMINAL OFFENSES BEING REPORTED TO THE POLICE. THE REPORTING PROPORTIONS FOR PARTICULAR CRIMES CAN CHANGE BECAUSE OF A CHANGE IN REPORTING PREFERENCE OR AS A RESULT OF A CHANGE IN THE CHARACTERISTICS OF CRIME WITHIN THAT CATEGORY. A MATHEMATICAL MODEL IS CONSTRUCTED TO DETERMINE WHETHER CHANGES IN THE PROPORTION OF VICTIMIZATION REPORTED TO THE AUTHORITIES PRODUCES SYSTEMATIC BIAS IN REPORTED CRIME ESTIMATES OF CHANGES IN ACTUAL VICTIMIZATIONS. IT IS FOUND THAT REPORTED CRIME COUNTS EITHER EXAGGERATE THE AMOUNT OF CHANGE IN VICTIMIZATIONS OR TEND TO MISREPRESENT THE DIRECTION OF CHANGE IN VICTIMIZATIONS. IT IS HIGHLY LIKELY THAT THE AMOUNT OF MEASUREMENT ERROR IN CHANGES IN REPORTED CRIME IS SIGNIFICANTLY LARGE IN COMPARISON WITH, IF NOT LARGER THAN, THE NORMAL AMOUNTS OF CHANGE IN REPORTED CRIME. IT IS CONCLUDED THAT FURTHER RESEARCH SHOULD BE DONE, USING THE VICTIMIZATION DATA, TO DETERMINE WHY THE OVERALL REPORTING PROPORTION INCREASED IN EACH OF THE 3 YEARS OF THE NCS, IN ADDITION, A GREATER USE OF VICTIMIZATION SURVEYS IS NEEDED. EVEN MORE IMPORTANT, ATTRIBUTE-BASED CRIME REPORTING AND THE DEVELOPMENT OF A BETTER CRIME-CLASSIFICATION SYSTEM SHOULD BE ENCOURAGED. TABULAR DATA ARE GIVEN AND AN APPENDIX PROVIDES AN ALGEBRAIC ANALYSIS TO PROVIDE IMPROVED UNDERSTANDING OF THE RELATIONSHIP BETWEEN VICTIMIZATION AND UCR-TYPE REPORTED CRIME. (AUTHOR ABSTRACT MODIFIED)