A series of five experiments was designed and administered to 679 elementary and junior high school students over a 2-year period to test the relevance of Jean Piaget's concept of conservation to musical learning. Musical tasks consisting of stimulus patterns and systematic variations of these patterns were designed for each experiment, and experimental settings for individual experiments were varied. Results indicated (1) task performance progressively improved from younger to older age groups; (2) improvement in conservation of tonal patterns preceded improvement in conservation of rhythm patterns; (3) training to enhance conservation was most effective at ages 5 and 7; (4) change of mode, contour, and rhythm pattern interfered with conservation more than change of instrument, tempo, or the addition of harmony; (5) a plateau in music conservation skills was reached in the fourth grade; (6) patterns in minor mode produced better rhythm conservation than major or atonal patterns; (7) the initial teaching of musical structure may best be pursued through the study of familiar music. The presence of visuals in one experiment made a significant difference in the results. The conclusions emphasize the importance of an early acquaintance with basic music structure and vocabulary and the need for instruction to decenter perception from the biasing aspects of music by a consideration of musical variations. (Author/LH)