A study was conducted with third-grade children in a university laboratory school to determine the relative effectiveness of a method for teaching recorder in which the sequential objectives are logically ordered by Edwin Gordon's empirical model of learning sequence and a traditional method which stresses note reading. The criteria used for comparison were change in the level of developmental music aptitude and performance achievement on the recorder. A review of the literature supported Edwin Gordon's theory of developmental music aptitude which proposes that until about age nine, environmental factors can affect the level of music aptitude, and, at about age nine, music aptitude stabilizes. The review of the literature also supported the use of singing activities with instrumental instruction, and the application of verbal association systems to tonal and rhythmic patterns as pedagogical techniques. Treatment for the control group involved a teaching-learning procedure in which individual fingerings, pitches, and rhythm symbols were presented in isolation and assembled in playing songs from notation. The method used with the experimental group involved a teaching-learning sequence in which children first learned to sing the song by rote. In learning to play the song on the recorder, each melodic and rhythmic pattern was isolated and initially sung or chanted. A verbal association system was then associated with the tonal and rhythmic pattern. The notation representing the pattern was introduced after extensive aural and verbal association experience, and after learning several songs. Primary sources of data included: pretest/posttest using Gordon's PMMA to measure developmental music aptitude change, and a rating scale test to measure recorder performance achievement. Results of the PMMA supported the alternative hypotheses that the experimental group had significantly higher mean composite and rhythmic increases than the control group. The mean increase in the tonal scores was greater for the experimental group, but not significantly higher at the .05 level. Results of the investigator-designed performance achievement tests supported the alternative hypotheses that the experimental method of teaching recorder was more effective in all dimensions--melodic, rhythmic, executive skills, and composite--than the traditional method.