In this study, the effectiveness of recorded models in the context of ensemble rehearsals was examined. Over a 5-week treatment period, directors of three middle/junior high and two high school bands systematically included professional recordings as part of their preparation of selected pieces. Students completed weekly self-evaluation reports assessing their individual progress and their ensembles' progress on modal and no-model pieces. Using numerical and free-response formats, students evaluated self-achievement and ensemble achievement on notes/rhythms, articulation/dynamics, tuning, and balance. Expert evaluations revealed no difference in achievement between model and no-model pieces on pretreatment and posttreatment performance recordings. Student evaluations showed more modest achievement gains for model pieces. High school students demonstrated more positive self-evaluations for their own versus their ensembles' performance and greater overall differentiation in their evaluations across time. Middle school/junior high students were significantly more positive toward the model pieces. All students provided a greater number of free-response comments for model pieces.