This study assesses the effect of mirrors on body image and performance in high and low performing female collegiate ballet students. Twenty-three females enrolled in a beginning ballet class were taught using mirrors, and a second group of 23 beginning females were taught without mirrors. All participants completed the Cash 69-item Body Self-Relations Questionnaire during the first and last class of a 14-week semester. They were videotaped performing in the studio during the fifth and fourteenth weeks. Two ballet teachers independently viewed the videotapes to evaluate the dancers' rhythmic accuracy, ease and flow of movement, and mastery of steps and alignment, and rated the students' skill level on a 1-5 scale. For analysis purposes, students whose scores averaged three or higher were categorized as "high performers," and those who averaged less than three were "low performers." Two (mirror, non-mirror) by two (high performance, low performance) by two (pre-test, post-test) repeated measures ANOVAs were used to test class differences over the course of the semester. There were significant 3-way interactions for overweight preoccupation (p < 0.01) and body-areas satisfaction (p < 0.05). Low performers increased in overweight preoccupation in the non-mirror class while decreasing in the mirror class. High performers significantly increased in satisfaction for most areas of their body in the non-mirror class, while there were smaller increases for both low and high performers in the mirror class. It is concluded that while use of the mirror has some benefits in training, higher performing dancers feel better about their body image when they do not use the mirror. Lower performers who use the mirror worry less about their weight; those who do not use the mirror worry more. The mirror may provide feedback that helps low performing dancers feel more comfortable with their weight.