STUDY COMMENCES WITH A REVIEW OF THE EXISTING TECHNIQUES FOR NAME IDENTIFICATION. LAW ENFORCEMENT AND CRIMINAL JUSTICE AGENCIES WERE ASKED FOR STUDIES THEY HAD MADE OR TECHNIQUES THEY WERE CURRENTLY USING. EVALUATION INDICATED THAT ALL EXISTING METHODS WERE DEFICIENT IN SOME RESPECT FOR USE IN A LARGE CRIMINAL NAME IDENTIFICATION SYSTEM. THESE EVALUATION RESULTS LED TO A SHIFT IN THE EMPHASIS OF THE STUDY TOWARDS THE DEVELOPMENT OF A NEW NAME SEARCH TECHNIQUE. THIS NEW NAME SEARCH METHOD - NYSIIS METHOD - OFFERS A VERY HIGH RELIABILITY OF GROUPING NAME VARIANTS TOGETHER WHILE RETAINING THE GROUPING CHARACTERISTICS REQUIRED FOR A VERY LARGE NAME SEARCH SYSTEM. THE CONCLUSIONS ARE THE NEW TECHNIQUE IS OPTIMAL FOR ALL LEVELS OF SYSTEM SIZE UP TO THREE MILLION ENTRIES. PRELIMINARY ANALYSIS INDICATES THAT VARIANTS OF THIS TECHNIQUE ARE APT TO BE OPTIMAL FOR LARGER FILES. THE NYSIIS METHOD IS SUFFICIENTLY COMPLEX TO RULE OUT MANUAL USE. IT IS SIMPLE ENOUGH TO PERMIT EASY IMPLEMENTATION ON MEDIUM TO LARGE SCALE COMPUTER SYSTEMS. SINCE MOST STATES WILL ULTIMATELY USE A COMPUTER SYSTEM FOR NAME IDENTIFICATION, THIS RESTRICTION SHOULD NOT BE A LIMITATION. IT IS RECOMMENDED THAT ANY STATE CONSIDERING THE DEVELOPMENT OF A NAME SEARCH SYSTEM USE THIS NEW METHOD IN THE DESIGN.