THE EVALUATING COMMITTEE INCLUDED EXPERTS ON ACOUSTICS, SPEECH SCIENCE, SPEECH PATHOLOGY, ELECTRONICS, ELECTRICAL ENGINEERING, AUDIO RECORDING SYSTEMS, AND CRIMINAL LAW AND LAWS OF EVIDENCE. THE ISSUE OF VOICEPRINT ADMISSIBILITY IN COURTS OF LAW WAS NOT ADDRESSED. THE COMMITTEE FOUND THAT (1) SOME INFORMATION ON A PERSON'S IDENTITY COULD BE OBTAINED THROUGH LISTENING AND LOOKING AT VOICEGRAMS; (2) VOICEGRAMS DIFFER FROM FINGERPRINTS BECAUSE THE SAME WORD SHOWS AN ACOUSTIC CHANGE, AT LEAST SLIGHTLY, EVERY TIME IT IS SPOKEN BY A PARTICULAR PERSON; AND (3) ERROR RATES OF AURAL-VISUAL VOICE IDENTIFICATION VARY WIDELY DEPENDING ON THE PROPERTIES OF VOICES, CONDITIONS UNDER WHICH VOICE SAMPLES WERE MADE, EQUIPMENT, AND EXAMINER SKILL AND KNOWLEDGE. THE COMMITTEE CONCLUDED THAT THE FULL DEVELOPMENT OF VOICE IDENTIFICATION BY BOTH AURAL-VISUAL AND AUTOMATED METHODS REQUIRED LONG-TERM RESEARCH AND DEVELOPMENT. THEY SUGGESTED THAT AN EXPLANATION OF THE TECHNIQUE'S LIMITATIONS SHOULD ACCOMPANY ANY PRESENTATION OF VOICEGRAM EVIDENCE. A BROAD NATIONAL PROGRAM OF SCIENTIFIC RESEARCH ON VOICE IDENTIFICATION AND OBJECTIVE STANDARDS AND METHODS FOR TESTING VOICE IDENTIFICATION EXAMINERS WERE RECOMMENDED. PHOTOGRAPHS AND FOOTNOTES ARE INCLUDED. (AOP)