The purpose of this study was to investigate the music preferences of elderly people, the relationship of elderly people to music, and their exposure to it. This study was a survey in the form of a questionnaire, and 501 elderly people over 60 years old took part. The results were analyzed by Crosstab Analysis to reveal differences according to four variables: gender, age, education level, and economic status. I. Musical Preferences of the elderly were found to be as follows: First, the preferred musical styles of the elderly were popular music, followed by Korean classical music, religious music, classical, and lastly western pop music. Second, the favorite music of elderly people were the songs they had enjoyed in their twenties and thirties. Third, the elderly preferred slow tempos. Fourth, the elderly preferred cheerful and light music the most. Fifth, the reason the elderly preferred certain music was because they liked the lyrics and mood of the music. Sixth, their favorite lyrics reminded them of past memories(27.9%), comforted their minds (25.1%), or made them feel nostalgic (23.6%). Seventh, the elderly didn\'t like the music of the younger generation. II. Responses to Questions about the Elderly and their relationship to music were as follows: First, 85 percent of the subjects liked music, while only 1.6 percent of showed no interest at all. Second, the elderly thought of music as \'very important\' or \'important\' in their lives. Third, in answer to the question concerning \'emotional help\', elderly people thought that music gave them great pleasure, or made them comfortable, or reduced tension and stress. III. Responses to questions about their exposure to music were as follows: First, most elderly people listen to music for less than one hour per day. Only 8.6 percent of them seldom listened to music. Second, their exposure to music came from TV, cassette tapes, radio, and records, in that order. The media choices revealed differences according to age and education level. Third, concerning number of recorded music purchases during a year, the response rating to \'No Purchase\' was the highest. Fourth, with regard to the number of cassette tapes purchased during a year, the response rating to \'1-3 tapes\' was the highest. Fifth, concerning concert attendance during a year, the response rating to \'No Attendance\' was the highest. Four variables, gender, age, and education level, influenced the results of this study while economic status did not appear to be a factor. The results of this study offer important, useful, and fundamental information for successful and efficient music therapy for elderly people. In particular, the differences determined by gender, age, and education level suggest that the individual backgrounds of clients should be considered in planning and practicing music therapy for elderly people.