SPEAKER RECOGNITION IS DEFINED AS ANY DECISION-MAKING PROCESS THAT USES THE SPEAKER-DEPENDENT FEATURES OF THE SPEECH SIGNAL. ORIGIN AND NATURE OF THESE FEATURES ARE DISCUSSED FIRST. THIS BACKGROUND MATERIAL IS FOLLOWED BY DETAILED DESCRIPTIONS OF THREE GENERAL METHODS OF SPEAKER RECOGNITION. EACH METHOD IS ILLUSTRATED BY MANY EXPERIMENTAL STUDIES. SPEAKER RECOGNITION BY LISTENING APPEARS TO BE THE MOST ACCURATE AND RELIABLE METHOD AT THE PRESENT TIME. SPEAKER RECOGNITION BY VISUAL COMPARISON OF SPECTROGRAMS IS USED IN THE FIELD OF CRIMINOLOGY, BUT THIS METHOD MUST BE STUDIED FURTHER TO DETERMINE ITS VALIDITY. SPEAKER RECOGNITION BY MACHINE IS LIMITED BY VARIOUS DESIGN SHORTCOMINGS. MORE RESEARCH ON THIS METHOD IS EXPECTED TO IMPROVE RECOGNITION PERFORMANCE. A LIST OF REFERENCES IS INCLUDED IN THE SURVEY (AUTHOR ABSTRACT)