Purpose: The upper trapezius to serratus anterior muscle activation ratio is essential for optimal shoulder function. An alteration of this ratio, specifically a decrease in upper trapezius and increase in serratus anterior activation, is a main area of focus in shoulder rehabilitation (Kibler, 1998; Paine & Voight, 1993). Electromyography (EMG) biofeedback has been shown to be an effective rehabilitation technique to address many musculoskeletal disorders but there is limited research on the retention of improvements seen with EMG biofeedback (Ma et al., 2011; Lim et al., 2014; Weon, et al., 2011). The purpose of this study was to determine if EMG biofeedback can be used to improve scapular control by decreasing the upper trapezius to serratus anterior activation ratio. A secondary purpose was to determine if these predicted improvements in the ratio can be retained beyond the timeframe in which the treatment is provided. Methods: Twenty college aged (age=21.75±1.77) subjects (10 males, 10 females) volunteered to participate in this study. Subjects were randomized to the exercise only group or EMG biofeedback group. The exercise only group performed three exercises twice a week for four weeks with supervision. The EMG biofeedback group performed the same exercises twice a week for four weeks with the addition of watching EMG biofeedback on a computer monitor with the instructions to decrease the upper trapezius activation and increase the serratus anterior activation by adjusting the corresponding lines on the monitor. The percent maximal voluntary contraction (MVC) for each muscle during each exercise was measured on visit one, visit nine (after the four weeks of practice) and visit ten (after a two-week retention period). The ratio and the individual muscle changes were analyzed using multi-factor ANOVAs against group, exercise, and group by exercise interaction. Results: There was no significant effect of any of the variables on the ratios visit one to visit nine, nor when comparing visit nine to visit ten. The was a significant effect of group on the upper trapezius when comparing visit one to visit nine (p=0.007) with no effect seen comparing visit nine to visit ten. There was also a significant effect of group on the serratus anterior activation for both visit one to visit nine (p=0.000) and visit nine to visit ten (p=0.001). Conclusion: EMG biofeedback did not decrease the upper trapezius to serratus anterior activation ratio, but the individual muscle activation changes indicate that EMG biofeedback is effective at altering muscle activation rates in individual muscles and that those changes can be retained beyond the timeframe of the intervention. Additional research is needed with more subjects and in populations with shoulder pathologies to further investigate the effectiveness of this concept.