This study explored the effects of singing on the mood of singers. Participants, a community sample of volunteers, were randomly assigned to either a singing (experimental) or a listening to singing (control) group. The singers participated in a half-hour session of singing while the listeners sat and listened to the singing group. The Profile of Mood States Questionnaire (P.O.M.S.) was administered immediately before and after the singing session and again one week later. Multivariate analyses of variance (MANOVA) (3 x 2 factorial with three levels of time and two levels of group) were conducted on each of the P.O.M.S. subscales. Multivariate F tests indicated that significant changes occurred on the P.O.M.S. sub-scales (tension, anger, fatigue, vigour and confusion) for both the singing and listening groups over time. There was no significant group-time interactions indicating that both groups responded in a similar fashion to the singing session, although the effects for singing were more robust. The results of this study indicate that both singing and listening to singing can alter mood immediately after participation in a short singing session, and that some of these effects were evident in the P.O.M.S. scores one week later. These results suggest that a longer and more vigorous singing session is needed to obtain additional benefits of singing over listening.