Self-perception theory suggests that premenstrual syndrome (PMS) may arise from the misattribution of hormone-induced bodily changes. If so, individual differences in the role of bodily responses in emotional feelings, measured in a separate expression-manipulation procedure, should be related to susceptibility to PMS. In Study 1, women responsive to cues from their bodies showed significant mood changes, both negative and positive, with their cycle, over a 60 day span; whereas women relatively unresponsive to personal, bodily cues showed no consistent cycle effects. PMS was also predicted by a measure of emotional complexity. In Study 2 women whose moods were based on bodily cues also rated their moods as less positive if they were in their premenstrual week, and women unresponsive to their bodies were unaffected by their cycle. A reminder of their cycle-stage prevented PMS in the body cue group, which is a kind of discounting effect. Women who were unresponsive to their bodies also did not show PMS, and were unaffected by the reminder. 1 The name Premenstrual Syndrome implies a more severe problem than most women encounter, and the term premenstrual tension would seem more appropriate, if a bit too narrow, because tension is only one symptom. However, common usage is to refer to cycle-related emotional changes as PMS and we will follow that custom.