Background and purpose:Many studies have explored the benefits of social robots in healthcare, but fewer have delved into their application and effects in mu-sic therapy groups. Nevertheless, this is an issue of great significance deserving fur-ther research. This study compares the therapeutic factors of music listening group activities led by social robots versus those led by medical staff in an elderly care center. Methods:This project employs a repeated measures design, building upon a pilot study involving 20 subjects aged above 65 years. These subjects underwent eight sessions of music listening group therapy under the supervision of social ro-bots, followed by another eight sessions supervised by medical staff. The variances in therapeutic factors between these two groups were examined using a therapeu-tic-factor questionnaire administered immediately after the fourth and eighth week. Results:In week 8, the questionnaire scores for the music listening therapy group sessions directed by robots were either higher or equal to those in week 4 for 10 out of the 12 therapeutic factors. Specifically, the factors "Altruism," "Group cohesive-ness," and "Imparting of information" reached statistically significant levels. Con-versely, the scores for all 12 therapeutic factors in the sessions led by medical staff were higher or equal to those in the robot-led sessions, with each factor reaching statistical significance. Conclusions: Social robots may have applications in music therapy groups; however, music listening group therapy sessions supervised by re-al-person medical staff are still superior to those conducted by social robots.