In this study, fifty-three high school student wind instrumentalists participated to investigate the effectiveness between using a recorded model and a live model during music instruction. The students were initially tested on sight reading ability and groups were constructed to be equally divided by sight reading skill. All students were recorded sight reading an etude at the beginning of the treatment period. Following this initial pretest, the students underwent five treatment sessions of twenty minutes each. A control group (N = 18) was given verbal instruction only along with being instructed to vocalize rhythms, sing, and play the assigned etude. The first experimental group (N = 18) had the same treatment conditions as the control along with the use of an ideal recording of the etude performed by a viola performance major from a local university. The second experimental group (N = 17) also had the same treatment conditions as the control. In addition to these conditions, the same violist was physically present with this group to perform the etude live as a treatment condition. At the conclusion of the treatment period, the students were recorded playing the etude again as a posttest. The students were also given a five-question survey to indicate their perception of the effectiveness of the instructional methods used.Four instrumental music educators were selected to evaluate all of the pretest and posttest recordings using a rubric. Their scores were averaged together for each individual assessment. The data from the pretests and posttests was examined to determine the growth of each individual student. Results indicated that all three groups had a significant gain from pretest to posttest. The highest growth was observed in the experimental group utilizing a recorded model. The group using the live model had lower growth than the control group. Student responses from the groups with modeling indicated that the modeling condition was the most effective teaching technique used in rehearsal.